PEER GRQ‘UP FREENDSHIPS EN ONE CLASS 0? HIGH SCHQOL GIRLS: CHANGE AND STABELETY Thesis foo: {Em Emma of MI. E). MICHEGAN STATE U’NWERSETY Heme? A113 Kefiey 1966- ”163‘ —-— ...‘...Alsf_ I Lynn any L] NliCiz.1 LC ’ University i\\\ W“ \\ \\ \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\iii 3 1293 10413 7710 This is to certify that the thesis entitled PEER GROUP FRIENDSHIPS IN ONE CLASS OF HIGH SCHOOL GIRLS: CHANGE AND STABILITY presented by Eleanor A. Kelley has been accepted towards fulfillment of the requirements for Ph . D degree in §OC1010gY :; Major professor Date October 5 #1966 0469. tr. ”A ~~.-C -1‘ BLUE Ct .' pa »8 ‘K‘ :5 -\n~:. v a . ~ V34“ .._. y ABSTRACT PEER GROUP FRIENDSHIPS IN ONE CLASS OF HIGH SCHOOL GIRLS: CHANGE AND STABILITY by Eleanor A. Kelley The purpose of this study was to investigate change and stability longitudinally in informal adolescent peer groups within the formal school organization, and the influence of family social class ranking in forming informal adolescent peer groups. An entire population of adolescent girls at a suburban Midwestern high school was traced from their ninth grade entrance in the school through their senior year. The majority were daughters of highly educated, geographically mobile white collar residents. However, a few were the daughters of less educated blue collar workers, creating a comparative situation in which to study the interaction, or lack of interaction, of social class extremes. Data were obtained from the responses to two instruments, a background questionnaire and an interview schedule. The adolescents were ranked according to the social class rank of their parents using Warner's Index of Status Characteristics. Peer group social acceptance, defined as a mutual awareness and desire for interaction among the respondents, was Operationally defined by using the near-sociometric information, the girls' best friend choices. A classification, includ- ing two categories of acceptance, or lack of acceptance, reciprocal friendship structures and isolates, was established. f",_;1, -..o-... .L.$Ac C-I~~:” ‘ I ‘ I v-evA-n-c Tau-pa... .:C._.J...C -x...:..-u.. . , 7 , ..-..-,.._> . .y .L-.i:.:.»\ SET; -5 .1 o -; “yrs I.“ , A t- - ‘- ‘ ~" ..:\:\n . ‘ a- -V IF- ~~= Sum: --. "" F v. - us. ‘v‘. ”fig-d ‘. -rcbe ~l , Llrj ‘.::SE Ray V «E‘ts a. t :5“: and ISl A.‘ v q‘ 3‘. u: C Eleanor A. Kelley While extensive change was evident at all levels of the structural analysis, four definable sub-system complexes emerged as they split and recombined longitudinally. The sub-systems showed a definite town and university split which reflected parental social class placement as follows: The "social elite" was a closed sub-system composed predominantly of girls whose fathers were stable resident town professionals. The ‘members ranked higher in social class, were most often listed as "most popular" and "best dressed," but were not the highest academic achievers. The second town closed sub-system, a polarity to the social elite, was composed of girls whose fathers were blue collar workers. These girls ranked low in social class, were often named as "not dressed right," and were low academic achievers. This sub-system contained five of the eight known school dropouts for the entire study. The third and fourth sub-systems contained the majority of girls whose parents were university affiliated and some town girls. Both Open sub-systems, with interaction between them as well as with mutual pairs and isolates, were closely parallel in characteristics. However, one enjoyed a little more prestige than the other, ranking somewhat higher in social class and academic achievement, and receiving a few popularity and dress choices. It also contained the structure which emerged as the "intellectual elite." A comparative analysis of the data from two social acceptance perspectives, aggregate numbers based exclusively on choice status and reciprocated structured groups, revealed that different response patterns \ s . n. ..- .1 ' 3: h" 5 Erin , -oq ,~ .’.__,-- ‘- Lg- - Kr -_ ' ~~ _,_ ..-- a .4. .~ ,.c 4“ F,, ”r 011‘ Ht q -A Ra-‘ “‘A.~ Eleanor A. Kelley emerged for the two types of questions. Responses to the friendship question which establishes an interaction framework in the respondents' minds were distributed throughout the four sub-system.complexes; whereas responses to the choice status questions concerning p0pu1arity and dress, which were not designed to elicit interaction answers, were directed to girls in selected subdsystems whether or not the respondent perceived herself as a member of the sub-system. The perceptive awareness by the reSpondents of the social class related group sub-systems was also evident in the unreciprocated choices. Respondents chose, or were chosen by, sub-system.members who reflected their own social class characteristics. To summarize, the study traced the deve10ping informal peer friendship structures of an entire class of high school girls from ninth through twelfth grades with their unique historical features undistorted by sociometric manipulation, and revealed that the structures continued to reflect parental social class placement. Furthermore, friendship interaction among the girls at a polarity in social class ranks was highly limited. PEER GROUP FRIENDSHIPS IN ONE CLASS OF HIGH SCHOOL GIRLS: CHANGE AND STABILITY by Eleanor Ann Kelley A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1966 As a portion e “jar of persons perscns are aC‘rleCI-‘l' Dr. Joanne E1 it: her contriburis 1: :3e preparation Dr. Alan Bees; assist ace in pier: is: of the manuscri The previous :crlec'ged in text 1 'u‘illians, Arlene B. PYE'ViOUS researchei t' H ; Drier ptiase of th s" v .3: Q ata analysis L LE Departme 0: Hflm we Economics Dr. V . ACKNOWLEDGMENTS As a portion of a longitudinal project, this research has involved a number of persons who contributed to its culmination. The following persons are acknowledged for their specific contributions: Dr. Joanne Eicher, the project initiator and my thesis advisor for her contributions throughout the study and her insightful suggestions in the preparation of the manuscript. Dr. Alan Beegle, Dr. Jay Artis, and Dr. Ruth Hill Useem for their assistance in planning my program of work and their constructive critic- ism of the manuscript. The previous researchers whose specific contributions are ack- nowledged in text footnotes. They include: Betty Wass, Madelyn Williams, Arlene Bjorngaard Ostermeier, and Suzanne Hendricks. Each previous researcher acknowledged specific individuals who contributed to her phase of the longitudinal study. Additional contributers to the data analysis include Hazel Ogilvee Baxter and Barbara Gibson. The Department of Clothing, Textiles, and Related Arts, College of Home Economics which conducted the longitudinal project, especially Dr. Mary Gephart,the department chairman,who enthusiastically supported the project. The Michigan State University Experiment Station which furnished the funds to conduct the project. ii .= ' l v F. '7‘ F..“C-ram, ' V ..-- v-A. "V‘ .-‘c any L. 5 Last, The high school which participated in the study, especially the principal, counselor, teachers, and the girls who graciously gave their time and opinions to the interviewers. Last, but far from least, my friends and family, especially my parents, who have continually encouraged and assisted me throughout the graduate program. iii www- ‘ l n-- 7 Oct. \ u-.- “b. a..- \ .O-. u. 1.3.3..er a~~~J._.‘ "7 Tait? ': “ ban-nag...— ‘ n ‘1. VIII”: '3': -‘ .-v~o_ 0.. '5 iDr. '6 on. -.-Q .Q' - "" ¢~§ '4 . ?v c.. ‘9‘ -‘.- La.k . Q .1. U 0 R022: State: COQCE Otis: "M- so.“ RETTEW - ~—H TABLE OF CONTENTS ACKNOWLEDGMENTS. LIST OF TABLES . . . . . LIST OF FIGURES. . . . . . . . . . . . . . . . . . . . LIST OF APPENDICES . . . . . . . . . . . . . . . . . CHAPTER I. II. III. INTRODUCTION. . . . . . . . . . . . . Statement of Problem. . . . . . . . . Conceptual Orientation. . . . . . . . Orientation of the Present Study. . . . . . . . Summary . . . . . . . . . . . . . . . REVIEW OF LITERATURE. . . . . . . . . . Grammar School Sociometric Relationships -- Hallworth . . . . . . . . . . . . . . . . . The Social System of the High School -- Gordon. Adolescent Society -- Coleman . . . . . . . . . Summary . . . . . . . . . . . . . . . . . DEFINITION OF TERMS AND STATEMENT OF HYPOTHESES . Definition of Terms . . . . . . . . . . . Hypotheses. . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . IV. METHODOLOGY . . . . . . . . . . . . . . . . . Selection of Community. . . . . . . . . . . . . POpulation Selection. . . . . . . . . . . . . . Variables . . . . . . . . . . . . . . . . . . . Instrument Development. . . . . . . . . . . . . Instrument Administration . . . . . . . . . . . Operational Definitions . . . . . . . . . . . . Reliability and Validity as Related to Socio- metry . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . iv Page ii vi ix xi 14 16 19 22 24 24 34 41 42 42 45 46 46 48 49 60 64 p0 ;'p'rg 5....A .- O E: a c O o u. 0—4 LOSGI'L'DIXEQ Pspuiatic chzitudi' Longitudi' Corpssz‘ Social CZ Ctange 1: Change ir In7ivid;a Relatzc hangs ir Summary. mscrssrm ceases c Descripti Locate: Descripti Locate: Summary c Longit; TVelffi Suzmary' SUMMARY, C CO. ’Y . 'I «rare. E GenEra} RECOjr—g cofltrib. CHAPTER Page V. LONGITUDINAL POPULATION DESCRIPTION. . . . . . . . . 67 Population Size. . . . . . . . . . . . . 67 Longitudinal Social Class Composition. . . . . . . 69 Longitudinal Social Acceptance Structural .Composition . . . . . . . . . . . . . . . . . . 72 VI. DISCUSSION OF FINDINGS: WHETHER AND WHERE POPULA- TION FRIENDSHIP CHANGES OCCURRED . . . . . . . . . 79 Social Class Composition . . . . . . . . 79 Change in Peer Group Friendship Patterns . . . . . 88 Change in Size of Social Acceptance Categories . . 100 Individual Choice Status vs. Reciprocated Relationships. . . . . . . . . . 102 Change in Membership Among Respondents . . . . . . 113 Summary. . . . . . . . . . . . . . . . . . . . . . 124 VII. DISCUSSION OF FINDINGS: HOW POPULATION FRIENDSHIP CHANGES OCCURRED . . . . . . . . . . . . . . . . . 127 Description of 1962 Ninth Grade Positions as Located in the 1965 Twelfth Grade. . . . . . . 127 Description of 1965 Twelfth Grade Positions as Located in the 1962 Ninth Grade. . . . . . . 136 Summary of Sociometric Patterns of the 105 Longitudinal POpulation Members: Ninth Through Twelfth Grades . . . . . . . . . . . . . . . . . 146 Summary. . . . . . . . . . . . . . . . . . . . . . 157 VIII. SUMMARY, CONCLUSIONS, RECOMMENDATIONS, AND CONTRIBUTIONS. . . . . . . . . . . . . . . . . . . 161 General Conclusions and Implications . . . . . . . 173 Recommendations for Future Research. . . . . . . . 179 Contributions of the Study . . . . . . . . . . . . 181 APPENDICES. . . . . . . . . . . . . . . . . . . . . . . . . 185 BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . . . . . 222 1. Fear year a chazze 1: :: senior alas: ... Four year 5 AF: 1 ‘n, Stelai C.a: SIZE SCECC. V Nut wear 7 1 ttrai C022. gcries rar- ciass, 19e'l . - Summary by 1962 min t .iris in or Samar), b‘.‘ 1963 team girls in Or ‘1' Sumary b;- 1913; E‘LE‘SE the girls “1' Sunzarg k \Sh3 13E One hi*- V K .w‘ '¢.I f . 0&1! \c : J Lrtefia 1... . SQ; CC L V 1X, 991‘ 1 SIZE C the E Table II. III. IV. VI. VII. VIII. IX. XI. LIST OF TABLES Four year numerical summary of the stability and change in the p0pulation of girls in one high school senior class, 1965. Four year summary by percentage and number of the social class distribution among the girls in one high school senior class, 1965. Four year numerical summary of the population struc- tural composition in the reciprocal friendship cate- gories formed by the girls in one high school senior class, 1965 . . . . . . . . Summary by reciprocal friendship structures of the 1962 ninth grade social class composition among the girls in one high school senior class, 1965 Summary by reciprocal friendship structures of the 1963 tenth grade social class composition among the girls in one high school senior class, 1965 Summary by reciprocal friendship structures of the 1964 eleventh grade social class composition among the girls in one high school senior class, 1965 Summary by reciprocal friendship structures of the 1965 twelfth grade social class composition in one high school senior class, 1965. Four year longitudinal continuity of the reciprocal friendship structures among the girls in one high school senior class, 1965. Four year numerical summary according to membership size of the reciprocal friendship structures among the girls in one high school senior class, 1965 .. Four year numerical summary of the structural pat- terns formed by the reciprocated friendships among the girls in one high school senior class, 1965 Four year summary of cohesion scores distributed by individual reciprocal friendship structures among the girls in one high school senior class, 1965. vi Page 68 71 77 81 82 84 86 89 93 95 97 D. “‘13. g3 i 1': i: 4 . Fear year 2: reciprocal amen: tne g1 Nuzerical c choice sza: graie reci; h1g2 schoc; Fear year s ility and c isolate ca: scnool sen; FOUP year 5 =1 t“re e: or aZOrg the “V ‘5‘ “#4 rt“.3383: H15: 52" FOgr 3M5. ‘ s -L duke see: ‘4 ‘In ta. ‘Ed t?“ V}. - Lye) Table XII. XIII. XIV. XVI. XVII. XVIII. XIX. Page Four year mean cohesion scores of the individual reciprocal friendship structure cohesion scores among the girls in one high school senior class, 1965. . . . . . . . . . . . . . . . . . . . . . . . 98 Four year summary by number and percentage of the size, or number of occupants, in the reciprocal friendship categories among the girls in one high school senior class, 1965 . . . . . . . . . . . . . 101 Numerical comparison of reciprocated choices and choice status among the girls in three ninth grade reciprocal friendship structures in one high school senior class, 1965. . . . . . . . . . . 108 Four year summary by number and percentage of stab- ility and change among the girls included in the isolate category membership position in one high school senior class, 1965. . . . . . . . . . . . . 114 Four year summary by number and percentage of the stability and change among the girls in recip- rocated dyads in one high school senior class, 1965. . . . . . . . . . . . . . . . . . . . . . . . 116 Four year summary by number and percentage of the stability and change among the girls included in a mutual pair membership position in one high school senior class, 1965 . . . . . . . . . . . . . 118 Numerical summary by reciprocal friendship struc- ture of the ninth grade dyadic membership positions among the girls in one high school class which became the senior class, 1965 . . . . . . . . . . . 120 Three year numerical summary of the girls continuing as reciprocated dyadic membership positions in the reciprocal friendship structures formed by one high school senior class, 1965. . . . . . . . . . . 122 Four year summary by RFS complexes of the recogniz- able sociometric patterns among the girls with con- tinued membership in one high school senior class, 1965. . . . . . . . . . . . . . . . . . . . . . ... 148 Four year summary of the shifting sociometric pat- terns among the girls with continued membership in one high school senior class, 1965. . . . . . . . . 151 vii "e n a" i b erical s~ 5;: cap ce C8 R u a in one hit: 9 s S 1'20? V N; 5| :e 55, l: tsL¥aL {is i c.ass, in o cla KITS,” .. ‘ Percentaze p‘.) Table 10. 11. Numerical social class distribution by social ac- ceptance categories among the 1962 ninth grade girls in one high school class which became the senior class, 1965. Numerical social class distribution by social ac- ceptance categories among the 1963 tenth grade girls in one high school class which became the senior class, 1965. Numerical social class distribution by social ac- ceptance categories among the 1964 eleventh grade girls in one high school class which became the senior class, 1965 Numerical social class distribution by social ac- ceptance categories among the 1965 twelfth grade girls in one high school senior class, 1965. Percentage distribution according to individual ISC indices and reciprocal friendship categories among the 1962 ninth grade girls in one high school class which became the senior class, 1965. Percentage distribution according to individual ISC indices and reciprocal friendship categories among the 1963 tenth grade girls in one high school class which became the senior class, 1965. Percentage distribution according to individual ISC indices and reciprocal friendship categories among the 1964 eleventh grade girls in one high school class which became the senior class, 1965. Percentage distribution according to individual ISC indices and reciprocal friendship categories among the 1965 twelfth grade girls in one high school senior class, 1965. . . . . . . . . . Individual sociometric pattern for the girls in one high school senior class, 1965. Numerical comparison of reciprocated choices and choice statuses among the girls in one 1962 ninth grade high school class which became the senior class, 1965. Numerical comparison of reciprocated choices and choice statuses among the girls in one 1965 twelfth grade high school senior class, 1965 viii Page 195 196 197 198 199 200 201 202 203 213 217 u \" I v '. . ..uh' . .Ill‘l' C 5:- v‘ ‘ \ . .. M-Li ’,-'{ _“Ul_.¢..-: . C§ . V W Simul‘ -) rho C 1 av Figure 10. LIST OF FIGURES RFS 7, the "Leading Crowd? among the ninth grade girls in one high school class which became the senior class of 1965. . . . . . . . . . . . . Four year summary of the evolving splits and re- combinations as the RFS 7 complex deve10ped among the girls in one high school class which became the senior class of 1965. Reciprocal friendships of the ninth grade girls in one high school class which became the senior class of 1965. Reciprocal friendships of the tenth grade girls in one high school class which became the senior class of 1965. . Reciprocal friendships of the eleventh grade girls in one high school class which became the senior class of 1965. Reciprocal friendships of the twelfth grade girls in one high school senior class of 1965 . Reciprocal friendships of the ninth grade girls in one high school class which became the senior class of 1965, with individual choice status in- dicated for each girl. . . . . . . . . . . . . Reciprocated friendships of the twelfth grade girls in one high school senior class of 1965, with individual choice status indicated for each girl. . Reciprocal friendships and appraisal of p0pu1ar- ity and dress of the ninth grade girls in one high school class which became the senior class of 1965. . . . . . . . Reciprocal friendships and appraisal of p0pu1ar- ity and dress of the twelfth grade girls in one high school class of 1965. . Page 54 55 73 74 75 76 103 104 106 .107 .. i‘VnV :s‘b‘be H H. Reciprocal one high sc' of 1963, 91' position in H. Reciprocal in one big? grade reci; each girl. Figure 11. 12. Page Reciprocal friendships of the ninth grade girls in one high school class which became the senior class of 1965, with twelfth grade reciprocal friendship position indicated for each girl. . . . . . . . . . 129 Reciprocal friendships of the twelfth grade girls in one high school senior class of 1965, with ninth grade reciprocal friendship position indicated for each girl. . . . . . . . . . . . . . . . . . . . . 137 I- .qq‘flflo .‘d.C..~.x .. H ”veg-iv: i 1 D-\‘ -- ' . u . . VS\U.\ 2 \'-'~"."* F in: ' h‘bh‘hs. ... 3. Tables l-ll LIST OF APPENDICES Appendix Page A. 1. BACKGROUND QUESTIONNAIRE. . . . . . . . . . . . 185 2. INTERVIEW SCHEDULE . . . . . . . . . . . . . . 189 B.. Tables 1-11. . . . . . .. . . . . . . . . . . . . . 194 xi The research ‘11. study of the c: 3' 32.125 'I'lc'llll the l 2311 facets of pee 521111 acceptance c. Every socie: 512316 generation; “m1 In a small uerQ’EIY through my Lexi than; ing s [:3 EZ‘ 31 norms so... \ 13%;;- .lAn“5eT o: A") MChapt Er V’l “irate“ LI CHAPTER I INTRODUCTION Statement of Problem The research presented in this paper is a four year longitud- inal study of the change in informal adolescent peer friendship groups within the formal school organization. The study investigated three facets of peer group changes, structural patterns, size of social acceptance categories, and membership positions. Conceptual Orientation Every society is concerned with self-perpetuation beyond a single generation; thus the socialization process which molds a new born individual into a social being is crucial for societal sur- vival.1 In a small, stable society this process can be accomplished adequately through informal adult-child associations, whereas in a complex, changing society, rapidly changing roles tend to make parental norms somewhat obsolete as models for their children. 1A number of sociologists have discussed this process, includ- ing: Kingsley Davis, Human Society (New York: The Macmillan Co., 1949), Chapter VIII, pp. 195-233; Kingsley Davis, "The Sociology of Parent Youth Conflict," American Sociological Review, V (1940), pp. 523-532; and Talcott Parsons, "Age and Sex in the Social Structure of the United States," American Sociological Review, VII (1942), pp. 604-611. " .. : ‘H .- '. .16 atCidledthf‘. gang to assume r 12:15 izereasingl to the 'acreasin; r2115. Highly 5 3.216 education aislescents at. thezagaified '1: 521111101 become Although triest youths t titttaat socia ~38 evolved, E: 2 The socialization process is especially relevant in training the young to assume roles functionally requisite for the societal main- tenance process. In this situation a formal educational system be- comes increasingly important especially in American society due to the increasingly complex training requirements of some work roles. Highly Specialized roles demand that youths remain longer in the educational system. When the facts of larger numbers of adolescents and longer training periods are considered together, the magnified importance of the school as a key socialization in- stitution becomes apparent. Although the school has assumed primary responsibility to orient youths to the larger society, the family still remains an important socializing agent. And a third extremely important agent has evolved, especially during the adolescent period. Informal age-peer group associations have been described as tied to both the family and the school.1 According to Parsons, peer associations perform two psycholog- ical functions. 0n the one hand, the peer gram)may be regarded as a field for the exercise of independence from adult control . . . . But another very important function is to provide the child a source of non-adult approval and acceptance.2 The importance of the peer group for socialization in Our type of society should be clear. The motivational 1Talcott Parsons, Social Structure and Personaligy (Glencoe, 111.:The Free Press, 1964), p. 138. 21bid., p. 139. fcuncations 51.3; $1315, a of a hierarc cf :1: 111': i: associati it: tion it "national hierarchical caper. 2:5. 121 pl’CCESE F‘ -Afl'fl “Lento! ...';:::y are sne 3 foundations of character are inevitably first laid down through identification with parents, who are generation- superiors, and the generation difference is a type example of a hierarchimfl.status difference. But an immense part of the individual's adult role performance will have to be in association with status-equals or near-equals. In this situation it is important to have a reorganization of the motivational structure so that the original dominance of the hierarchiafl.axis is modified to strengthen the egalitarian components. The peer group plays a prominent part in this process. Elementary school peer groups, characterized by boundary fluidity and sharp sex segregation, are tied more closely to the child's resident neighborhood.2 As the adolescent period emerges, associations become more diverse. The child is exposed to a wider range of age- peers because the school structure shifts due to increased size, wider geographic origins of the school population, and diverse class partic- ipation resulting from elective subjects. Due to the shifting structure, Parsons' impression is that "the transitions to Junior High School and Senior High School are apt to mean a considerable reshuffling of friend- ships."3 Researchers have studied friendship patterns effectively through sociometric techniques. In fact, "the result of the sociometric develop- ment has been that the investigation of the smallest social aggregates has become more interesting than the larger ones; and that pint-sized revolutions, for instance social change produced in the classroom, 1Ibid., p. 139- 140. 21bid. 3Ibid., p. 150. I «A ‘ A W ’A' :31: :eLcue thins: .. in Q a w ‘ F‘fi fi‘s aaLaL‘LLa v ' F t...’.. , ...... y‘-V. .4..-E. duk- .dLC b v t I ‘ " fifi‘fiw n 0' o: d ~:.~...... I :--' ' it-.. . . " “ w e . \ ”.5. 51.6..Le .. .oov-.,. .,.J . “'“Vidvc‘ . Iv--.,,.- ~ u r'svbt3b _t ..nfi;r ‘1 "“. 5.1.g. . Y‘nb. assty.‘cn t 7'; . e~€ 1&2: 4 have become more interesting than efforts at a world wide revolu- tion."1 Sociologists recognize well the importance of social change. MacIver and Page noted that "society exists only as a time sequence. It is a becoming, not a being; a process, not a product."2 When social change is referred to as a process, the idea of continuity is introduced: A process means continuous change taking place in a definite manner through the operation of forces present from the first within the situation. Thus we speak of the group process, or the manner in which the relations of the members of a group, once brought together, acquire a certain dis- tinctive character. . . . In studying a process we ob- serve a series of transitions between one state of being and another. There is no necessary implication as to the relative quality of the two states of being, or as to the direction followed.3 Numerous change processes occur simultaneously in every society.4 The latter idea is supported by Parsons, Bales,and Shils. Us- ing a social system scheme, they stated: The system operates through the interaction of its member units. Every change or state in one unit, 112-: in its location in relation to any or all of the dimensions, in its energy change, etc. will affect all the other units in the systems and in turn the effects of these on the other units will "feed back" to the original unit.5 VIII (1945), p. 118. 1J. L. Moreno, "Sociometry, Comtism and Marxism," Sociometry. 2Robert MacIver and Charles Page, Society: An Introductory Analysis (New York: Holt, Rinehart and Winston, 1949), p. 511. 31bid., p. 522. 4Ibid., p. 523. 5Talcott Parsons, Robert F. Bales,and Edward A. Shils, Working Papers in the Theory of Action (Glencoe, 111.: The Free Press, 1953), p. 167. EVSCEZS and the; '. .;‘.‘:",“fi 9 ,_ 1v».b.\.il '\. ....- c . IIOQ-r-fiOI a- "5.....t.rn..a . I r.- ~..'.u 3,...r.'_ ...“. ...:st. 2" ":‘j’ a .- ... ..- .-t . v v n . \‘r- V ‘P “it. n u I q..; ,,'..: A» ."“" '“sha: are .. ..- \ Q g "4.. ‘1'“! n e . fl"; . . -'._3:< C" . .,_ Q 'n, I -u51.~v-a ‘“'..~‘“; 5‘ 5» y s We. 1 ~5th‘Kl." .-. ..\| H- ::::e-\ .- "‘S 1*; a.“ EES;“‘. v 525 1» 1“4 ‘. 21.52,: *0 37' . 13,5. I "r’ 1.: thrg 5 The change processes occur in a temporal order, therefore the systems and their unit processes must be treated as changing over time. In addition to differentiating the processes which change, the researcher must differentiate the patterns which remain relatively constant through time.1 Differentiation is also necessary between total system change and change in the units comprising the system. "Phases of system state are. . . to be regarded as resultants of phase changes in the units which are also systems."2 The action schema from which the above ideas have been drawn "can be extended from the microscopic study of small groups. . . for short periods to large scale phenomena over longer periods of time."3 Simmel also recognized that group interaction occurs at varied societal levels. In his discussion of the isolated individual, the dyad, the triad, and complex structures, he suggested that "our re- lationships thus develop upon the basis of reciprocal knowledge and this knowledge upon the basis of the actual relations. Both are in- extricably interwoven."4 Placing the interaction process in the larger societal context be emphasized, "If society is conceived as inter- action among individuals, the description of the forms of this inter- action is the task of the science of society in its strictest and most essential sense."5 11bid. 21bid., p. 168. 31bid., p. 170. 4Georg Simmel, The Sociology_of Georg Simmel, Trans. and ed. K. H. Wblff (Glencoe, I11.: The Free Press, 1950), p. 309. 51bid., pp. 21-22. ' v ;’ t "7' '. a a 523::C, sd 5“ I" \‘fia n ...-n “ ch :oDV rUbb 1L. 1: ye have had I ‘ . ' 0" 0a. . ,. ‘ s 0' .‘u .4 .L . q . . . " ' "NH—s --.. *z-cb‘.ul~ 3"». ' v n; ”NW-:3“ . —.. _ .M‘I ' ‘ o s. v-u-p :: H ‘ :VC“ “ .e C 1&213231 5:; s v . «’21.? v a: E ' l " .4 1 .- ‘a‘es SCCLC. 9356 Stent. subjm to < :- n. I ‘ Or _ ‘ A 1...: hole 11'221t: u . ru' 11 .4‘ . M . ine r -, 16:. '£:-.; A 1 - "‘“fie r s... A ‘.‘ l I N ..p: ‘Z . I‘ a ' a~.,q c. .A .\ VCV:5e ""5: in- ‘v ‘ Co in" I ‘::‘|:3 Y L“ ' I ’ ‘33. . 9 3. i“. :'. N‘K: \: xr‘cr. 6 Although sociologists recognize the importance of dynamic social change, they often neglect it in sociological research. Yet, as Homans stated, "The study of dynamics opens a whole new range of problems; it also puts in our power a more convincing method of exposition than any we have had up to now, though we may not be able to use the method for all it is worth."1 Perhaps one of the major problems inherent in studying social change can be found in C. W. Mills' discussion of the Marxian concept of "historical specificity:" Even if we are concerned with some limited area of one national social structure--we need historical materials. Only by an act of abstraction that unnecessarily vio- lates social reality can we try to freeze some knife- edge moment. . . . Knowing that what we are studying is subject to change, on the simplest descriptive levels, we must ask: What are the salient trends? To answer that question we must make a statement of "from what" and "to what”.2 Thus the longitudinal study will "freeze a series of knife-edge moments"as a means of establishing the sc0pe of "'from what' and 'to what'." Orientation of the Present Study The reported longitudinal investigation studied adolescent ac- ceptance in informal peer friendship groups within the formal school organization from two perspectives, social class and social acceptance.3 1George Homans, The Human Group (New York: Harcourt, Brace and Company, 1950), p. 335. 2C. Wright Mills, The Sociological Imagination (New York: Grove Press, Inc., 1961), p. 151. élgfigg, pp. 32-33 for definitions of social class and social acceptance. Q 7 ‘ “OI’10-4 4‘ £5~E..LC\. adtae . . ’ ..... ,. .... :v¥.:. 5.635 S» en‘s: ~ 4— " ... bar‘sab .A' “c.“ I 1 ‘ ‘ us: any: . r «— u».~3.v€s65.\.u s..'. .‘ ‘ibugrs 1:, u.scu55 :A:“la‘ ?' ". "~c ‘ 15.,ag‘z .. 5.. w' . ‘ H .. up An “2': -“ V Unneci tr ?e .' t N 7 The first objective evolved from one factor which is considered in— fluential in adolescents' friendship choices. Sociologists have char- acterized adolescent peer groups as coinciding with their parents' social class stratification rankings.1 This characterization was empirically demonstrated by Coleman: Social class was found to be a consideration in status evaluation; however, its importance did vary from school to school.2 Warner stated that insufficient at— tention has been given to social class interrelations.3 The first 1Barber stated that "adolescence is a period of intensive socialization, a period of learning to be an adult in the fashion de- fined by one's social class." Bernard Barber, Social Stratification (New York: Harcourt, Brace and World, Inc., 1957), p. 281. Other authors discussing the relationship of adolescent peer groups to family social stratification patterns include: Joseph A. Kahl, The American Class Structure (New York: Rinehart and Company, Inc., 1960), p. 135 and Allison Davis, "Socialization and Adolescent Personality," Read- ings in Social Psychology (New York: Henry Holt and Co., 1952), pp. 520-531. Research studies supporting this idea include: Mapheus Smith, "Some Factors in Friendship Selections of High School Students," Sociometry, VII (1944), pp. 303-313 and August B. Hollingshead, Elm: town's Youth (New York: Wiley, 1947), pp. 212-215. 2Social class seemed to be an important value governing a number of the respondents' choices. It was a criterion influencing membership in the "leading crowd." While social class emphasis varied from school to school, there seemed to be a direct variation according to the proportion of the upper social class students. Support for the "privileged class" theory of leadership seemed present in all except one school. James Coleman, Adolescent Society (Glencoe, 111-: The Free Press, 1961), pp. 103-110. 3W. Lloyd Warner, "The Study of Social Stratification," Review of Sociology, Joseph B. Gittler, ed. (New York: John Wiley and Sons, 1957), p. 236. . s .o ..- finale .4 :vou: V v r) :1 .a. d m ~n .. ...' SCCCQI s .4.- : u I‘ “'6 Va: ...»..cug a“ . a "..H." EEPh—‘ln. -...5 y L“_“‘_-‘ t ' ' w s. ‘;Q ‘Ha ,,F"\.‘ ..V, ‘“\‘ “Hens ‘¢A\ qe.,._‘_ ‘ ...:.<‘.‘.Llcn :1 "1' .. . ‘ .-.s at‘icv 2:2; V: I! 0" A ' DAA' eagle t g.."“ Vu. ‘ L" n A“ o :3! “‘kul: ..- e‘x' 8 study objective, therefore, was to investigate the possible influence social class has on the friendship choices, as one measure of social acceptance, among one population of high school girls over a four— year period. The second study objective concerns peer group social acceptance per se. Previous studies considered adolescent peer group acceptance in the school situation; but their mode of establishing peer group acceptance has been inconsistent, using diverse versions of the socio— metric techniques and yielding diverse findings.l Many earlier stud— ies, including longitudinal studies, incorporated sociometric technique variations which measured individual's acceptance in the total group yet did not include the reciprocity of relationships or patterns of interaction within the group.2 They are therefore not relevant for this study which adheres more closely to the assumptions underlying "true" sociometry. Basic to sociometry is the assumption of interpersonal relation- ships. Within this broad assumption, a number of requirements are necessary to administer a "true" sociometric test: 1The technique is "employed today in almost all studies of group functioning, very often in accompaniment with whatever other techniques particularly interest an investigator." Michael S. Olmsted, The Small Group (New York: Random House, 1959), p. 97. 2The Cannon longitudinal investigation used this type technique. Students were asked to name thier friends. Each individual was given a rank in the c1aSs based exclusively on number of times chosen. This rank did not include reciprocity of relations, iyep, the returned choice to those who chose him, or the creation of ties. Kenneth L. Cannon, "Stability of Sociometric Scores of High School Students," Journal of Educational Research, LII, 2 (October, 1958), pp. 43-48. 1) The 2) The c..oi 3) The Spec .) Te31 U! \_’ (I) I" O 6) The sad Sociometric tea tear-sociometrj L fat: 9.391.115 3:”. kph“ of res "1A. V 3 4 18::tiy 9 o 9 1) The limits of the group should be indicated to the subjects. 2) The subjects should be permitted an unlimited number of choices or rejections. 3) The choices or rejections should be made in terms of specific criteria. 4) Test results should be used to restructure the group. 5) Choices and rejections should be made privately. 6) The questions should be gauged to the subjects' level of understanding. Sociometric tests which fail to meet all six requirements are termed near-sociometric.1 Relatively few of the total given were true socio- metric tests. Often the criterion of restructuring the group is omitted. In fact, Bjernstedt found in investigating 100 sociometric studies that only 11 percent met the criterion of restructuring the group.2 Yet many school-focused studies of associational patterns had the social therapy, or group restructuring, purpose. In fact, Coleman stated that associational patterns in the school situation without the as- sumption of restructuring the group have been studied only infre- quently.3’4 A? 1Gardner Lindzey and Edgar F. Borgatta, "Sociometric Measure- ment," Handbook of Social Psychology, ed. Gardner Lindzey (Reading, Mass.:Addison-Wesley Publishing Company, Inc., 1954), p. 407. 2Ake Bjernstedt, Interpretations of Sociometric Choice Status (Lund, Sweden: CWK Gleerup, 1956), p. 49. 3Ibid. 4Many school oriented studies focused on elementary or kinder- garten groups, or situations in which a wide age range was present. Moreno, Jennings, Neugarten, and Northway are a few sociometrists conducting classroom oriented studies. An often cited social therapy Associationa 52:5: ezbracing :11. :1rizec' their izpc: E‘sese associat; cut others, pg; ies and away f; Perso‘s’ weake. SCCCI'eoftne N Brookover a: 1::io: patterns apj. iieas, they conclc. :13: but varying 1 cf :hazge than sta 'u'ider use c: \ .11? Was COLrCuCre Emaifling SCEC 2:12.335 LeaCErs‘v ‘0“t1’38t. Cook :3 school Situa t :. xic Study ofa \ X (19“5) l COIEzan’ r —J 10 Associational patterns constituted one phase of the Coleman study embracing nine high school adolescent social systems. He sum— marized their importance: These associations help reinforce certain values; under- cut others, pull energies in the direction of some activit— ies and away from others; strengthen the prestige of some persons, weaken that of others. They are, in part, a source of the culture, and in part are determined by it.1 Brookover and his associates concluded from an associational pattern study with a longitudinal perspective that variations in inter- action patterns appeared.2 Using the cited study to support their ideas, they concluded that, while many longitudinal studies have cited high but varying levels of stability, "there is decidedly more evidence of change than stability in the structures." [of the cited study].3 Wider use of longitudinal methods, following a particular group through repeated measurements has been recommended.4 Thus the second study was conducted by Jennings using a wide age range sample of New York Training School girls, a relatively "closed" situation. Helen Jennings, Leadership in Isolation (New York: Longmans, Green Co., 1950). In contrast, Cook used the social therapy technique in an "Open" high school situation. Lloyd Allen Cook, "An Experimental Socio- graphic Study of a Stratified 10th Grade Class," American Sociological Review, X (1945), pp. 250—261. ‘1Coleman, op. cit., p. 173. 2Wilbur B. Brookover, Orden C. Smucker and John Fred Thaden, Sociologygof Education (New York: American Book Company, 1955), p. 213. 3Ibid., p. 218. Bracket insert is this writer's addition. l'Carter V. Good and Douglas E. Scates, Methods of Research, (New York: Appleton-Century-Crofts, 1954), p. 805. . 3... ~.;:" cbiectix‘ an: 0'39 PCP'J (B Several 1) Pee 2) The 3) The 11 study objective was to investigate the peer group friendship changes among one p0pulation of high school girls over a four year time period. Several assumptions underlie the study design: 1) Peer groups exert an important influence on adolescent be- havior. 2) The sociometric test can be used to index an individual's peer group acceptance or rejection. 3) The sociogram can be used graphically to present the at- tractions in a sociometrically tested group. 4) Adolescents share their families' social class position. Summary The importance of adolescent peer groups in the socialization process has been established. Adolescent acceptance in informal peer groups within the formal school organization was the tepic of numerous research projects, each using many sociometric technique variations. Associational patterns, including reciprocated relation- ships, were infrequently the research focus; longitudinal associational pattern studies were even more infrequent. Yet associational patterns are an important aspect of the informal adolescent social system in the school, and longitudinal methods are useful to obtain a picture of group change or stability. Thus, the major focus of this study was to consider group interaction patterns and membership change and stability longitudinally, utilizing a near-sociometric question to establish friendship choices in one p0pulation of adolescent girls. .. r .C {a D; it 9?. it I u .... I S . i C .t 2: C ..C n u A.“ V“ .‘¢ «xv C .2 C .l C 3 b u 9:5 ... C; .u. nu. .... Vflb .QA Nthu r ... v. w“ .0. ... .4.‘ . b . a a . .. 12 The reporting of this study will follow the following pattern: Chapter II, Review of Literature; Chapter III, Definition of Terms and Hypotheses DevelOpment; Chapter IV, Methodology; Chapter V, Des- cription of Population; Chapter VI, Discussion of Findings Pertain- ing to Whether and Where the Groups Changed; Chapter VII, Discussion of Findings Pertaining to How the Groups Changed; Chapter VIII, Sum- mary, Conclusions, Implications, and Recommendations for Future Research. .3 vs 1e::e IVES l .-.- -- . . D‘ -- ‘fl. . “3 _‘_;--o ' .. Chl- .. E588 an“. a . a ‘.. ‘1- e as V vs ti Q ~~uplt t 'v I CHAPTER II REVIEW OF LITERATURE The two objectives of the study were the investigation of l) the influence of family social class ranking in the formulation of peer group friendships and 2) stability and change in adolescent peer friendship patterns. The literature has been investigated with these objectives in mind. Researchers interested in American adolescent behavior gener- ally accept the idea that adolescent subcultures do exist.1 Yet the methods used to study their subcultural behavior have been varied in design and the findings, while indicating high stability of group acceptance patterns, tend to vary over time.2 The majority of classroom studies have included sociometric technique variations not directly applicable to this study.3 In fact, a literature investigation revealed only three studies which directly focused on associational patterns in a natural situation without the assumption of using the research findings to restructure the group. The three studies are: Hallworth's "Sociometric 1David Gottleib and Jon Reeves, Adolescent Behavior in Urban Areas (East Lansing, Michigan: Bureau of Research and Publications, Michigan State University, College of Education and Cooperative Extension, 1962), pp. III-4. 2Supra, p.10, footnote 2. 3Supra, pp. 7-8 for the variations cited. 13 _ ~-. . fi,_'..‘ :;E-5’.l-¢‘:“l $554151 - - r x,‘ I‘ b V ‘— 3,.‘ (7,-01’8 a c «333) ..d 14 Relationships Among Grammar School Boys and Girls Between the Ages of 11 and 16 Years,"1 Gordon's The Social System of the High School,2 and Coleman's Adolescent Sociegy.3 The first is a journal article encompassing one year's findings from a three year longitudinal study; the others are monographs coverning only one year. Grammar School Sociometric Relationships——Hallworth The Hallworth study was conducted in a suburban London, England: coeducational grammar school between June, 1947, and June, 1950. Respondents included three hundred students in eleven forms (classes); however, the journal article reported case studies of only four forms (one each, first, second, third, and fourth) which contained 150 subjects and spanned only one year. Seven hypotheses relating to group structure were formulated and the general summary statement concerning them noted that "substantiation of these hypotheses would imply that sociometric rank is a function not only of personality but also of group structure and development, and that although constant over a short period of time, it would be subject to change over longer periods."4 Sitting together, working together,and friendship were the basic criteria for the sociometric tests. Choices were limited to 1H. J. Hallworth, "Sociometric Relationships Among Grammar School Boys and Girls Between the Ages of 11 and 16 Years," Sociometry, XVI (1953), pp. 39-700 2Wayne C. Gordon, The Social System of the High School (Glencoe, Ill.:The Free Press, 1957). 3James Coleman, The Adolescent Society (Glencoe, Ills:The Free Press, 1961). 4Hallworth, op. cit., p. 40. . o 1 o -‘A‘ ’zree h'ltfiln I ... :xxluded that ‘ reszited from t': conclusions whit 1. "The for: wEre used. 2. ”The Posi: PEare OVer- fix U0 :1» H H Ln H H (7* r ,. ’ J' H‘ ‘3‘ m. - I U ’U 0‘ r-4 7‘ (— H- L O L 15 three within-form choices and rejections were allowed but not re- quired. Case study discussions of the various girls' and boys' form- groups disclosed differential structural patterning and change. Hallworth concluded that the hypotheses relating to group structure and development "were generally substantiated by the findings which resulted from the sociometric case studies of the groups."1 Specific conclusions which seem currently relevant are: l. "The existence of relatively stable groups inside each form was shown by means of sociograms, and these groups were at any one time much the same on the three criteria used."2 "The distribution of sociometric choices was almost positively skewed, and the values of the groups ap- peared to be centered upon those individuals who were over-chosen on the sociometric tests."3 "There was a tendency to make progressively larger and better-integrated groups."4 ". . . When attention had been centred upon nuclear group and upon one individual in particular, the nuclear group continued to attract still more choices whilst the one individual received fewer."5 "All of the large, stable groups reported above did, in fact, form slowly over a period of one year or more, and persisted for a further period of one or two years. There were cases, however, in which development took place much more rapidly in response to a specific stimulus from the environment. In these instances the group was centred around values which were of immediate but less permanent significance to its members, and the group it- self generally had a shorter life history."6 11bid., p. 69. 21bid. 31bid., p. 68. 4 Ibid. 51bid., p. 69. 61bid. are dev preserv structu: This : ifipassiag the O. western Suburb 16 A limiting factor, from the viewpoint of a possible current com- parison, is the wide age range (11 through 16 years) in the study— forms. Hallworth stated additional study limitations: Wider investigation would be necessary to establish whether the generalizations drawn from the study of this limited number of children's groups can be considered to apply also to other groups. Further, the acceptance of these hy- potheses would raise many other questions, concerning, for example, the variations of the normal process of development, the conditions under which secondary modes or organization are develop, and the means whereby group integration may be preserved without the formation of too rigid a hierarchical structure.1 The Social System of the High School--Gordon This monograph discussed the findings from a case study en- compassing the social system formed by 576 students in a single Mid- western suburban high school. Specifically, the study was designed to explore "the theory that the dominant motivation of the high school student is to achieve and maintain a general social status within the organization of the school."2 The research focused on three sub-systems within the school organization as a composite general social status: "1. The formal organization of the school which prescribes learning achievement; 11bid., p. 70. 2Gordon, Op. cit., p. l. 7 The 5: extra 3. he n fri :2 The netwc 'fifi’l‘yn .....-gh the res; :etric question were separated are directly 21;: The nint :. I“. . .ected bv be ESF‘itations 1 ef ‘::Egrat10n SEE ie.ative s tatus a' .eits,2 In the Sailing" , ‘- ‘ orleut a T"1 ‘.. pr 17 2. The sytem of student organizations usually referred to as extracurricular activities; and 3. The network of interpersonal relationships defined by friendship choices."1 Only the third, the network of interpersonal relationships, applies directly to the study under discussion. The network of interpersonal relationships was determined through the respondents' school-wide choices in response to a socio- metric question. Both boys and girls were included, but sex and grade were separated in the analysis. Only the findings regarding girls are directly apropos to this discussion. The ninth grade girls had relatively low group orientation as reflected by both out—of—school and up-grade choices. High mobility aspirations left large numbers of unchosen individuals. Influence integration seemed to be around individuals rather than cliques. Relative status equality among the cliques minimized individual clique influence. Girls were chosen as friends; boys as date ob- jects.2 In the tenth grade group, the girls exhibited the highest in- school orientation. Apparently there was a low cross-sex interest. Two prominent cliques, a "social elite" and an "intellectual elite" 1Ibid., p. 3. 21bid., p. 14. l merged as group-. :lcsure accoupani: Eleventh g: of cliques rather flexed value 6.1;” 1 53365. Excess: :igns but, as 0335.53)t 8:1:e amup inter; :53; 9 ates. IntEns ‘ ‘uenCe re: l " 1' : “ n ’ u ‘“‘ «q; 35 . all frlefid ‘ ‘ .' 1 sn‘- 'a' 'U 18 emerged as group-dominating. An increased tendency for structural closure accompanied the increased clique status awareness.1 Eleventh grade group influence was competitive among a number of cliques rather than dominated by two. The dominant cliques re- flected value differences, and the tendency for closed structures emerged. Excessive clique closure stemming from competition left many unchosen members. In-school orientations were still relatively high and cross-sex friendship choices continued low. Cross—sex interest was high, but as romantic idealization rather than friendship choices.2 Cross-sex interest was primarily instrumental in low twelfth grade group integration. Out-of-school friendships also reflected the breakup of the social system. Absence of potential up-grade choices led to more in—group choices. Mutual pairs absorbed potential isolates. Intense honors competition led to broken clique solidarity and influence reintegration around personal leaders.3 To summarize, Gordon found that group interaction patterns varied both between grades and within grades. A strength of the study is the inclusion of all four grades, or the entire school p0pu- lation. Especially currently relevant is the finding that 85 percent of all friendship choices were made toward members of the same grade and sex group; thus "these groups each constituted virtually lIbid., p. 16. 21bid., p. 18. 31bid., p. 20. iziependent 521 school."1 2225 urrent 1 I g . . . ...e 5139 ¢ ;. .. ; ouLlLtla.‘ C 19 independent sub-systems of informal organization within the school."1 Two of Gordon's general implications are pertinent to ques- tions currently under consideration: are Clique structures have remarkable stability and most of the dominant cliques were those which have been formed in grade school as early as the 6th grade. Nevertheless, modifications took place; old members dropped out and new members were incorporated. The size and structure of small groups seem to be a function of their position in the prestige structure, and therefore, significantly related to the social class position of the members' families in the community.3 In addition, Gordon's comments on longitudinal investigation apropos: The variations in group structure as represented in the profiles manifest an organization-disorganization cycle in the process of group development. Since each group has certain unique features in its history affecting its develOpment, we are not able to determine the exact varia- tions in the confi urations of the structure without a time- sequence analysis. Adolescent Society--Coleman Coleman obtained fruitful research suggestions from the above studies.5 However, his monograph reported a study which was much 139191., p. 80. 2_I_b_1_g_., p. 106. 3314., p. 134. 4;2;g., p. 81. 5Coleman, op. cit., p. 335. 20 larger in sc0pe. The study included the entire value systems of ten high schools' social systems.1 Only a very small portion of the mono- graph, the discussion of associational patterns, is currently apropos and, unfortunately, limited in value because the discussion was largely confined to smaller school associational patterns. The larger, urban school situations, more parallel to the current research setting, were treated in other discussion facets, but very little in discussing associational patterns. The smallest sociometric unit appearing on Coleman's sociogram was a two person mutual choice structure. Thus, the omission of isolates and unreciprocated choices constitutes a limitation. But special focus on cliques with a basic nucleus of four or five as well as focus on the larger structures is a strength.2 Boys' and girls' cliques were recorded separately, and Coleman noted that the girls' clique structures were far more complex than the boys' structures. Both sexes exhibited differential complexity and membership between the large and small schools and between grades in each school, ipg., there seemed to be more mutual pairing and a lower proportion of total group members in definite cliques in the large schools than in the small schools. Also, large school complex struc- tural patterns differed. Between two of the large schools, Coleman 1A tenth school was briefly mentioned by Coleman; however, only nine were discussed extensively. Ibid., p. 56. 2Ibid., p. 174 and p. 183. a'tributed this c aies of the scho; class bases . .ne sch-col displayed :rzaaity functi. Tne remai 5"3‘3'86 around I Sihcels see: It social class I: The lefidin the snide: ative Of 1 high-Educ; more li‘lfie “he col Oriented activitie cliques I aria“ ted c l 1, 21 attributed this difference to factors such as: the differential ages of the schools, community solidarity, and differential social class bases. The parents in the predominantly upper middle class school displayed greater school interest and participated in more community functions and after school activities.1 The remaining associational patterns in Coleman's discussion evolved around the smaller schools; yet a few reflections regarding differential structural dominance patterns between the smaller schools seem fruitful for the current study. The influence of family social class ranking was recognized by Coleman: The leading crowd must be in some fashion "in touch with” the student body as a whole, although not fully represent- ative of it. The higher the proportion of white-collar, high-educationa1 background students in the school, the more likely that the leading clique will be a predominantly white collar one--which in these schools means one more oriented to adult goals, a college education, and to school activities and interests. Yet these white-collar dominant cliques may not be particularly interested in learning.2 A second set of factors influencing the emergence of a middle class oriented clique was the less tangible adult community actions, the school administration, and the teachers. This emergence was due largely to middle class domination of the school program through influence gained in organizational participation in activities such as the PTA.3 11bid., p. 185. 21bid., p. 215. 31bid. .I ‘--§_—fl In sum, 2 praportion of w‘: tlass clique do iet status upse system. Cole: vere non-schoo school oriente led to peer r1 335 fostered oriented gr 0‘ 22 In sum, a direct relationship seemed to exist between the pr0portion of white-collar families in a community and the middle class clique domination within the adolescent in-school groups. Yet status upsets were apparent in some grades within two school systems. Coleman explained the upsets: some middle class cliques were non-school oriented, while the working class cliques were school oriented. Teacher and administrative recognition and rewards led to peer recognition and prestige for the lower class cliques and fostered a situation in which school oriented and non-school oriented groups shared an unstable systemic dominance.l Summary Adolescent behavior researchers generally accept the idea that adolescent subcultures exist in American society. Yet the numerous adolescent behavior studies have used diverse methodological consider- ations and yielded inconsistent findings. The majority utilized sociometric technique variations which do not coincide with the cur- rent technique focusing on associational patterns derived from reciproc- ated relations. Therefore, only three studies which are directly applicable have been discussed. Neither monograph included a longitud- inal perspective and, while the journal article reported research which was conducted with a longitudinal perspective, only a one-year time span is reported. All three samples were wider in scape than the 1Ibid., p. 217. "- .... current pogulatic range. The threc the association tions from sch: Ear current me analISIS 0f bc ations, high . ranking on pe itudin31 res: grade Change that Some cf fer"511-11211 E In ‘ toiflilOwif a an 5.. ‘ Gila 1 findings 1 Passible Beta-Se d d 23 current p0pu1ation; the journal report also had a wider age- range. The three research projects found grade—to-grade variation in the associational patterns, and the Coleman monograph reported varia- tions from school to school. Findings which seem especially crucial for current methodological considerations include: the separate analysis of boys' and girls' groups, low-cross-sex friendship orient- ations, high within-grade choices, the influence of social class ranking on peer group formation, and the recognized need for a long- itudinal research design to enable accurate comparisons of between- grade changes in structural patterns. The researchers recognized that some cited changes in a single time focus could be due to dif- ferential group characteristics rather than age-grade changing. In view of the above findings, the present study scope, limited to following longitudinally one class of girls in one high school, is a managmflfle research design. The longitudinal emphasis is a special strength of the research design. Specific points regarding research findings from the three reviewed articles were elaborated to give possible reference points for comparison in the subsequent chapters. Because diverse sociometric approaches exist, the next chapter will define the terms and state the hypotheses which have been developed to guide this research. The term is as diverse a atseptance. ’1“; tarts and disc raided the res Intri< different pa tems are us 353501131 re‘ 3.5 an aggrel CHAPTER III DEFINITION OF TERMS AND STATEMENT OF HYPOTHESES The terminology used in adolescent peer group acceptance research is as diverse as the sociometric techniques used to measure peer group acceptance. Therefore, this chapter will clarify current research terms and discuss the development and statement of the hypotheses which guided the research. Definition of Terms Intricate group friendship interactions can be tied together in different patterns, structures or configurations. In fact, the three terms are used interchangeably in the literature to describe inter- personal reciprocated relationships. Moreno described a configuration as an aggregate of individuals; or consisting of two, or multiple, ways of interaction. "They are social wholes."1 His definition will be modified and extended to embrace not only the term configuration but also the terms patterns and structure and used in the following discussions. The modification is necessary to distinguish between a 1J. L. Moreno, Sociometry Reader (Glencoe, 111: Free Press, 1960), pp. 17 and 19. 24 stmctured grow? be initially amen. a ceilectivity ‘6 based. upon share hole“ which for individuals eng avareness of s‘: In his 3 aLso have stud to provide a 1 Group dynamic med the Va: grouPS-Z Tr. nap and the "‘3“? 63m @1113“ he t1 25 structured group and a simple aggregate of individuals who may not be mutually amenable to engaging in interaction. An aggregate may be a collectivity without mutual awareness forming a structured group based upon shared norms and behavior. For this research a "social whole" which forms a structured group is considered to be formed by individuals engaged in mutually accepted interaction based upon an awareness of shared norms and behavior. In his Sociometry Reader, Moreno noted that group dynamicists also have studied structural patterns and "two newer volumes set out to provide a much needed link between sociometry and group dynamics."1 Group dynamicists have presented relationship patterns graphically and named the various configurations; but they studied small experimental groups.2 The most frequent experimental group size seems to be five men and the researcher is usually task oriented. Generally the group dynamicists concluded that groups with the most complete inter- action networks perform most efficiently due to a high cohesion factor. 1Moreno, op. cit., p. 716. 2Bavelas and his colleagues are credited with setting the ground work for small experimental group research. In fact, an interesting analogy was drawn between the communication patterns in a circle and a wheel and larger societal patterns. The circle has been related to the "open", or democratic society and the wheel to the "closed", or authoritarian society. Michael S. Olmsted, The Small Group (New York: Random House, 1959), p. 104. One could conclude that insights gained in studying small groups might be related usefully to the larger society. 3mg describini chain, star, S"?- ieavitt‘ s sigzing the diii if two patte breaking a l cise definit a complex to He also 1 by ccmunicatio cases, if they sanitation whi A more the Leavitt d: 26 Terms describing the various configurations include: Y, Wheel, circle, chain, star, slash, comcon, and all-channel.l Leavitt's often cited study stated that his criterion for de- signing the different patterns was: if two patterns cannot be bent into the same shape without breaking a link, they are different patterns. A more pre- cise definition of unique patterns would require the use of a complex topological concept.2 He also stated that members of a group may be linked together by communication networks in numerous ways. "It is enough, in some cases, if they are each touched by some part of a network of com- munication which also touches each of the others at some point."3 A more complete presentation of unique structural patterns than the Leavitt discussion related the group interaction patterns to 1Ideas incorporated in presenting configurational patterns graphically may be found in: D. Krech, R. Crutchfield and E. Ballachy, Individual in Society (New York: McGraw-Hill Book Company, Inc., 1962), pp. 468-471; and Olmsted, op. cit., pp. 102-104. Specific research studies whose research designs used the communication nets, or con- figurational patterns include: Marvin Shaw, Gerard Rothschild and John Strickland, "Decision Processes in Communication Nets Upon Organiza- tion and Performance in Task Oriented Groups," Management Science, I (1954), pp. 233-250; Harold J. Leavitt, "Some Effects of Certain Com-, munication Patterns on Group Performance," Readings in Social Psychologx, Eleanor E. Maccaby, Theodore H. Newcomb and Eugene L. Hartley, eds. (New York: Holt, Rinehart and Winston, Inc., 1958), pp. 546-563. Be- cause most of the previous research describing communication nets has been conducted with 5-man groups, the described configurational pat- terns are limited to patterns within the scope of S-man interactions. 2Ibid., p. 39. 31bid., p. 38. 27 mathematical graph theory concepts.1i For example, the term lipg' in graph theory seems to be related to the term.pipyor‘reciprocated relationship used in the present research. Graph theory seems to offer a wide scope of knowledge which may be utilized to define terms des— cribing interpersonal relationships. The terms describing the various structural relations in this study were derived from ideas gleaned from various sources reporting small group research. Yet the operational definitions are unique to the study, having been developed for the larger project.2 Williams used a number of the terms first developed for the larger project,3 however, the current research includes some revised definitions and newly defined terms. The terms constitute a social acceptance classi- fication containing categories of the different reciprocated relation- ships, or lack of reciprocation. Basic to the idea of reciprocated relationships is the concept dyad which refers to reciprocated relations between two individuals.4 Examples of a two person group include friendship pairs, marital pairs, . 1Frank Haray and Robert 2. Norman, Graph Theory as a Mathematical Modelgin Social Science (Ann Arbor, Mich::University of Michigan, 1953). 2Joanne B. Eicher and Eleanor Kelley, "Adolescent Girls'View- points From Ninth Through Twelfth Grades Concerning Dress, Social Ac- ceptance,and Related Factors," Michigan State Experiment Station project #743, research in progress. 3Madelyn Williams, "Opinions on Clothing, Appearance and Social Acceptance As Factors in Group Cohesion of Ninth Grade Girls" (Unpub- lished Master's thesis, Michigan State University, 1963). 4The concept stems from Simmel's group relationships dis- cussions. See the previously cited reference. Simmel, op. cit., p. 309. business partner relationship wi terns dyad, or person reciproc or in an exclus trace the shif‘ 10 SEparate tw The ter recL‘Pfocated I SOD groUPS.2 9191 than the the term Yeti vg'a_A “‘-~ “Rowing t)": 1 . u 2. H 28 business partners, and parent-child dyads.l An exclusive two person relationship will be defined in this research as a mutual pair; the terms dyad, or dyadic relationship, will be used to describe any two person reciprocated relationship either within a complex structure or in an exclusive pair. This distinction is necessary in order to trace the shifting ties present in the longitudinal population and to separate two person exclusive dyads from any dyad. The term clique traditionally describes a configuration with reciprocated relations or ties, usually limited to four or five per— son groups.2 Because so many of the current structures were more com- plex than the usual clique pattern, a unique term has been developed, the term reciprocal friendship structure, abbreviated as RFS, which is defined as a sociometric diagram of individuals whose choices of friends were returned. Reciprocal friendship structures include the following types: 1. Mutual Pair: An exclusive two person reciprocated dyad.3 2. Triangle: Three individuals whose choice of each other forms a cohesive unit. 3. Chain: An open series of reciprocated choices. 1Howard Becker and Ruth H. Useem, "Sociological Analysis of the Dyad," American Sociological Review, VII (1942), p. 13. 2 See the Coleman clique definition which essentially utilizes this criterion. James Coleman, The Adolescent Sociepy (Glencoe, Illa The Free Press, 1961), p. 183. 3The researchers recognize that an exclusive dyad contains only one chance for group continuity. In other words, members of a group which contains three or more persons have more Opportunity for group continuity; whereas loss of one person in a mutually exclusive dyad leads to group disintegration. For this reason, the mutual pairs will be treated in the following chapters as a separate category rather than discussed within the framework of the RFS category per se. ‘i’ w ——W" .f-v" 4. W389 5. Squa 6. Con: Because traced as they category. A cc population can: or individuals segzent charac: . etc function i: isolation part. tribe them . A: friendships , l~ Iso 2- Iso 3. Iso 4. ISO 1?» 29 4. Wheel: Two chains which have a central common member. 5. Square: Four individuals Whose basic arrangement forms a square; additional internal ties may criss-cross between corners. 6. Complex structures: Consist of a cluster of dyadic re- lationships which may be arranged in a number of different patterns; including all, or some, of patterns 1 through 5. Because the entire group's associational patterns are to be traced as they change, it is necessary to consider the isolates as a category. A complete picture of the associational patterns in a population cannot be presented without including unchosen individuals, or individuals who do not choose, who represent a specific p0pulation segment characterized by a lack of ties. They are still group members and function in a specific manner, ippp, a state of isolation. Various isolation patterns may emerge; thus categories were developed to des- cribe them. An isolate, defined as an individual who had no reciprocated friendships, may be one of four types: 1. Isolate 1: A "true" isolate, or an individual who made no choices and received none. 2. Isolate 2 An "ignored" isolate, or an individual who made choices but received none. 3. Isolate 3: A "self" isolate, or an individual who made no choices but received some. 4. Isolate 4: A "confused" isolate, or an individual who made choices and received some, but the choices made and received do not match.1 1Williams, 0 . cit., p. 33. In additi gories, terns a‘. iaiiviinal posit l . Accur 4 - Con A Seco #- . c5101 Ce Status 516 g B .. fi‘: ~“ 30 In addition to the terms deve10ped to describe structural cate- gories, terms also have been developed to describe various types of individual positions in the structures. They are: 1. Accurate perceiver: An individual who had all choices reciprocated, or who neither chose nor was chosen. 2. Pivot: An individual who had four or more ties to other RFS members. 3. Peripheral member: An individual who was drawn into the RFS by a single tie and thus had a fringe position in relation to the integrated dyadic clusters. 4. Connector: An individual who linked two segments of a larger RFS or had a single tie to both cohesive seg- ments. A second type of individual measure was developed for the research. Choice status describes the individuals' ranking on the sociometric questions when reciprocity is excluded. Choice status was operationally defined as the total number of choices received by an individual whether or not she reciprocated them. The writings of group dynamic oriented researchers were studied for ideas concerning the concept cohesion which also must be defined. An article by Schachter and his associates divided the concept into two orientations: cohesiveness as measured by the morals, efficiency or spirit of the group and cohesiveness as measured by the attractive- ness of the group.1 In the same publication, Cartwright and Zander 1S. Schachter, N. Ellertson, D. McBride and G. Gregory, "An Experimental Study of Cohesiveness and Productivity," Group Dynamics, Research and Theory, D. Cartwright and A. Zander, eds. (Evanston, Ill: Row Petersen and Co., 1960), p. 152. — rw'w 7' cited a frequen and his associa' at all the iorc~ Taere may be di‘ L5, it may be need, or due tc individual to individuals ma attraction ma. Variou specific {8% tive all-Elbe: a COE'ParisQ‘ 0f out‘g‘CO‘c min base tetal POSS the same i .4 r 31 cited a frequently quoted definition of cohesion. developed by Festinger and his associates: "The cohesiveness of a group is the resultant of all the forces acting on all the members to remain in the group."1 There may be diverse sources for individual attraction to a group, 3:2,, it may be due to the group itself as the object of the individual's need, or due to the fact that being in a group may be a means for the individual to satisfy needs outside the group. Because different individuals may have different needs, the individual sources of group attraction may vary.2 Various cohesion measures have been developed according to specific research interests. They include: a comparison of the rela— tive number of remarks in which the respondents used "I" vs."We", a comparison of number of within-group friendship ties with the number of out-group friendship ties to the larger community, a friendship index based on a ratio of number of actual within—group selections to total possible number, determining the degree to which members share the same behavior and belief norms and standards, and an ingenious picture-projective test.3 1Darwin Cartwright and Alvin Zander, "Introduction: Group Co- hesiveness," Group Dynamics, Research and Theory, D. Cartwright and A. Zander, eds. (Evanston, 111: Row Petersen and Co., 1960), p. 74. 2Ibid., pp. 74 and 76. 3;b1d., pp. 70-71. None of research becaus cohesion measur tapt was oper at 532.50]. C indie and R the 2112:: everyone but S’ePresents iniivic'ual t SIC-up Choige Ember Of 3 formula in- 02in} An mums Of 31311:; at rs. "tn win‘s‘ 32 None of the above measures is suited entirely to the current research because they do not include reciprocated relations; thus a cohesion.measure including reciprocity was developed. The con- X where the N (N-l) symbol C indicates cohesion, X the total number of in-group choices, cept was operationally defined as follows: C and N the number of individuals in the group. An individual can choose everyone but himself; thus each individual can make N-l choices where N represents the group and -1 represents self; N—l choices for each individual times N individuals in the group gives the total number group choices possible if everyone chose everyone else. X is the total number of actual group choices, accounting for both the within-group unreciprocated choices and the reciprocated within-group choices. This formula includes reciprocity which many other social acceptance measures omit.1 An additional measure was developed to split the cohesive seg- ments of the larger RFS's for more detailed study: the segments were split at the point of a connector.2 In addition to the reciprocated within-segment choices, the within-segment unreciprocated choices were considered. They were used to determine the direction to place the 1The formula first appeared in: Suzanne Hendricks, "Opinions on Clothing and Appearance as Related to Group and Non-group Member- ship of Twelfth Grade Girls" (Unpublished Masterh thesis, Michigan State University, 1965), p. 110. 2For a definition of a connector see supra, p. 30. tmector, i_.§_ L.et nnreciproca were used to d». it the unrecip: the cluster we. sent was suffic also were util Social dEEISEd using Ham-er, CO:?] Passess some PEOPIQ in h]- SOCial QLaS in "igTzEr c Soc V's. Cue mlltua; 33 connector, ipg,, the connector was placed in the segment into which her unreciprocated choices were made. The unreciprocated choices also were used to determine how many segmental splits should be made,.ipg., if the unreciprocated choices were predominantly to a given cluster, the cluster was considered to be a cohesive segment. Provided the seg- ment was sufficiently cohesive to contain pivotal individuals,1 they also were utilized to determine central sections of a segment.2 Social class, one of the major variables, was operationally defined using Warner's Index of Status Characteristics. According to Warner, complex societies which service large populations always possess some kind of status system "which by its own values places people in higher or lower positions."3 Therefore, for this study, social class is defined as the status system which orders individuals in higher or lower positions in their social structure. Social acceptance, the second major variable, is based upon the mutual choice among group members to engage in interpersonal re- lationships stemming from an awareness of shared norms and behavior. From this vieWpoint, social acceptance is more than a simple aggregate or collectivity of individuals; social acceptance depends upon the 1For a definition of a pivot see su ra, p. 30. 2Hendricks, op. cit., p.45. 3W. Lloyd Warner, Marcia Meeker, and Kenneth Eells, Social Class in America (Chicago: Science Research Associates, Inc., 1949), p. 8. :reation oi i the Raisins i:ti its: 1 r aggregates ID; .‘ 4.x»:agiy ‘ - n ’1” .4 . 5"“.8 tr“ 34 creation of interpersonal structures with communication routes among the individuals filling specific positions in the structures.1 The individual positions create links which coordinate, or tie together, aggregates into structured groups.2 Social acceptance was oper- ationally defined through the sociometric configurations deve10ped from the girls' friendship choices, ipg,, the dyadic relations either within a mutual pair, or clustered in RFSs, and the isolates who represent lack of reciprocation. The operational definitions will be elaborated in the following chapter after the hypotheses developed to guide the research are presented. Hypotheses In keeping with the idea presented by Parsons, Bales, and Shils that the effect of change is reciprocated between the various unit levels in a social system,3 a series of hypotheses were developed to guide the investigation of change at the various structural levels. Underlying the hypotheses development was the attempt to account for the relatively high correlations cited in many longitudinal peer group 1Ideas leading to the definition of social acceptance were gleaned from a structural approach to considering small groups found in: Donald Olmsted, Social Groups, Roles, and Leadership (East Lansing, Michigan: Institute for Community Development and Services, Michigan State University, 1961), pp. 12-16. 2For some specifically defined group positions used in this research, supra, pp. 29-30. 3Supra, p. 4 for a discussion of their ideas. 35 associational studies. Possibly the correlations have been high be- cause a total group acceptance measure was obtained from individual choice scores excluding reciprocity of relationships. Thus an in— dividual can have a relatively stable sociometric choice score over time, but the choices may be entirely different individuals;1 or individual scores could change so that a total acceptance score for the entire group could be deceptive, i433, the increased acceptance of one individual could balance the decreased acceptance of another.2 However, when reciprocity is considered, the group structures could exhibit marked changes, while the group acceptance measure obtained by an aggregate total of choices could remain highly stable. Perhaps Brookover and his associates were referring to something of this nature when they stated "clearly there is more evidence of change than stability."3 1A closely related idea is: "As we have stated elsewhere, al- though a great deal of value has come out of the development of a scoring system in sociometry, the interest which has resulted in socio- metric status has often obscured the essence of sociometry, namely that its focus is on relationships. 'Scores which are statistically identi- cal are rarely sociometrically equivalent'." Mary Northway, "A Plan for Sociometric Studies in a Longitudinal Programme of Research in Child Development," Sociometry, XVII (1954), p. 273. 2The Cannon study contains this discrepancy. He stated, "In the cases of some individuals, rather marked changes occurred in their scores." Yet, describing some representative types of sociometric statuses, two scores appeared which seem to balance one another in pattern, ipg,, student E ranked 4th in the 9th grade and drOpped to 11th in the 12th grade. "The only ones who chose him in his last year were two buddies and his girl friend." Student I?was cited as an example of increased acceptance rank over the four year study, but no Specific rank change was given for her. Kenneth Cannon, "Stability of Sociometric Scores of High School Students," Journal of Educational Re- search, L11, 2 (October, 1958). 3Brookover, Smucker, and Thaden, op. cit., p. 218. S Ct? '. 5 uCaC' In C This re CC L».' l \§’C' as ens in i L. .5 . t?" ‘..e a. 9 u“ 36 This research emphasizes structure; therefore, individual respond- ents are considered only as they fill places in a category in the social acceptance classification. Two distinctions in category placement are made. In the number hypothesis the individuals are considered only as occupants creating the size of a specific social acceptance cate- gory. In other words, position within the category is ignored. The second dist\inction, necessary in the membership hypothesis, concerns individual position, or location, in the structures. The positions in informal structures do not remain consistent, having automatic replacement when one person leaves the system, or sub-system; nor do new persons entering the system, or sub-system, automatically assume a previously vacated position; nor must a position be vacated before a new person may assume a position in the informal structures. For these reasons, the membership hypothesis was developed to explore shifting structural positions as the configurations evolved over the four years. The assumption was made that, in an emergent informal open system in which adolescents are characterized as search- 1 ing for a self identity, there would be a continued "jockey- ing for positions," rather than jockeying only in the early stages of group development. In sum, the individuals are l The ideas concerning adolescent searching for identity were drawn from two sources as follows: Irene M. Josselyn, "The Older Adolescent," Values and Ideals of American Youth, ed. Eli Ginsberg (New York: Columbia University Press, 1961), pp. 31 and 32. Murray Wax, "Themes in Cosmetics and Grooming," American Journal of Sociology, IXII (1957), p. 591. ccnsidered ing. the population To Surat. 212‘: r to DEC; acceptance clé are a category cateEOry, a, I? is also C9151; may Chafige 10 Therefore the as nU—ZCEI . 37 considered impersonally as members filling structural positions in the population configurations,1 not as specific individuals. To summarize, membership refers to structural position and number to occupants creating the size of a category in the social acceptance classification, disregarding position within the category. The RFSs, composed of members with reciprocated friendship choices, are a category in which membership and number may change. The isolate category, although comprised of individuals with no structural ties, is also considered as a category in which membership as well as size may change longitudinally and thereby affect the p0pulation positions. Therefore the isolates are included in discussing membership as well as number. The following hypotheses were formulated from the overall hypothesis that structural patterns will change more when membership reciprocity is considered and change less when only the size, or aggregate number, in a given category is considered excluding reciproc- ity leading to specific positional placement. In addition, other hypotheses were formulated relating to previous research. Where liter— ature exists to support the hypothesis, it is cited. The major hypotheses are presented in the following order which does not infer a hierarchical ordering of importance: 1Some specifically defined positions in this study include connector, pivot and peripheral member. Su ra, pp. 29—30. 1. Parental eH BEtYlC “73% T' .u ' ‘ [he fifln‘.| “' A... V‘w‘... I'a: I4‘I‘ ‘ ““ R ‘ Sbbge'a“: fl 38 I. Parental social class rank will influence an adolescent's socio— metricgposition. A. There will be more RFSs‘with a homogeneamssocial class composition than RFSs with a mixed social class com- position;1 and B. of the mixed social class RFSs, the differences will be "variations around the mean" rather than sharply dif- .ferentiated social class rankings.2 II. Thegpopulations[#peer group friendship patterns will exhibit more change than stability over a fourgyear time span, and they will change as follows: A. Some RFSs will completely disappear between the 1962 ninth and 1965 twelfth grades, and B. some new RFSs will appear between the 1962 ninth and 1965 twelfth grades; therefore, C. few RFSs will remain in the 1965 twelfth grade which were RFS's in the 1962 ninth grade. D. The RFSs will become larger between the 1962 ninth and the 1965 twelfth grades,3 therefore, 1According to the cited stratification literature, adolescents tend to stratify in peer groups according to their family's social class rank. See Kahl, Allison Davis, Barber, Mapheus Smith, and Hol- lingshead, loc. cit. 2This idea was gleaned from Lloyd Warner's ideas: ”For purposes of clarity, the lines on the chart depicting the place of each class indicate sharp divisions. Actually, each class merges into the class above and the one below it. A class system where there is movement up and down by individuals and families in an open social system where there is territorial as well as social movement necessarily makes no sharp distinctions between one class and contiguous ones." W. Lloyd Warner, American Life Dream and Reality (Chicago: The University of Chicago Press, 1962), p. 79. The idea is especially relevant in a highly geographically mobile population like the current population. 3These hypotheses were derived from Hallworth's finding that his groups tended to become larger, fewer in number, and more inte- grated over the three year time span. Hallworth, "Sociometric Relation- ships Among Grammar School Boys and Girls Between the Ages of 11 and 16 Years," Sociometry, XVI (1953). : a S S C hA 39 E. fewer.RFSs will.be.present in the 1965 twelfth grade than in the 1962 ninth grade.1 F. The RFSs configurational patterns will change between the 1962 ninth and 1965 twelfth grades; therefore, G. few configurational patterns will be present in the 1965 twelfth grade which were present in the 1962 ninth grade. H. The individual RFSs will become more cohesive from the 1962 ninth through the 1965 twelfth grades;2 therefore, I. the RFS category will be more cohesive in the 1965 twelfth grade than in the 1962 ninth grade. J. Each year, the smaller RFSs will be more cohesive than the larger RFSs.3 III. The number of respondents in each reciprocal friendship cate— gorygwill exhibit little change over a four_year time span4 as follows: 11bid. 2Ibid. 3This hypothesis was derived from a statement made by Kretch and associates that group size affects the functioning of a group, $;g,, "Thus group cohesiveness and member satisfaction tend to be greater in smaller groups; there is evidence that larger groups inhibit par- ticipation of some members; style of leadership varies with group size; the relation between group size and effectiveness is complex and var- iable." Kretch, Crutchfield and Ballachy, Individual in Society (New York: MCGTaW-Hill Book Company, Inc., 1962), p. 470. Note the con- tradiction this poses to the last segment of the Hallworth statement (nipage 33- If the Kretch statement is true, then the hypothesis formulated from Hallworth's statement will not be supported. Recogniz— ing this, the hypothesis was still formulated according to Hallworth's finding since it was the only empirical study available for a guide- line. However, the Hallworth study is recognized as a case study dis- cussion excluding statistical analysis. l'This hypothesis was formulated following Gordon's findings that the number of respondents in cliques and the number of isolates remained stable with the exception of a breakdown of cliques in the eleventh grade. In order to deVelop general hypotheses for all four years, little change each year was hypothesized. This will not pre— vent a checking for parallelism with Gordon's eleventh grade find- ings. Supra, p. 18 for Gordon's findings. 40 A. There will be little change in the total number of isolate, mutual pair, and RFS members each year; therefore, B. there will be little change in the number of isolate, mutual pair, and RFS members in the 1965 twelfth grade from the number in the 1962 ninth grade. IV. The individual respondents' choice status excluding;reciprocated choices will not coincide withgposition in a structured group based upon reciprocated choices1 in two areas as follows: A. There will be a difference in individual choice status as measured by friendship choices received and population position as measured by reciprocated friendship choices, and B. there will be a difference in individual choice status as measured by popularity and dress choices received and population position as measured by reciprocated friendship choices. V. The membership in each reciprocal friendship category will ex- hibit more change than stability_over a four year timeperiod2 as follows: A. There will be more change than stability in isolate member- ship each year; therefore, 1The choice status vs. reciprocated relationships hypothesis is a direct outgrowth of the overall hypothesis underlying the development of the hypotheses. Su ra, p. 34. Furthermore, the researchers per- ceived in the interview situation an interaction definition of the friendship role by the girls as exhibited in their responses. A state- ment by Lazersfeld and Barton reflects the researcher's position con- cerning classification with an interaction framework. "The classifica- tion should present as clearly as possible the respondent's own definition of the situation--his focus of attention, his categories of thought." Lazersfeld and Barton, "Qualitative Measurement in the Social Sciences: Classification, Typologies and Indices." The Policy Sciences, Recent Developments in ScOpe and Method, Lenner and Lasswell eds. ‘ (Stanford, Calif: Stanford University Press, 1951), p. 157. 2The rationale underlying the development of the membership hypothesis is incorporated in the discussion defining the term mem- bership. Supra, p. 36. 41 B. there will be few isolate members in 1965 twelfth grade who were also isolates in 1962 ninth grade. C. There will be more new dyads created each year than dyads remaining stable from the previous year; therefore, D. there will be few dyads remaining in 1965 twelfth grade which were dyads in 1962 ninth grade. B. There will be more mutual pairs which exhibit membership change each year; therefore, F. there will be more change than stability in mutual pair membership in 1965 twelfth grade from mutual pair member- ship in 1962 ninth grade. G. There will be more internal shifts than stability of ties among the members who have reciprocated choices within the individual RFSs each year; and, H. there will be more internal shifts than stability of un- reciprocated choices within the individual RFSs each year; and, I. there will be more shifting memberships from one RFS to another than membership stability in the same RFS's each year, occurring at two levels: 1) single individuals and 2) dyadic clusters. Summary This chapter discussed and defined the unique terms developed to guide the research and presented the working hypotheses developed to guide the research. The next chapter will elaborate the methodology utilized Operationally to define the major variables, social class and social acceptance, and will discuss the research setting and p0pulation. t0 lfl'fes sols CE. ...-... J. ) the «J t. f. e . 4 55 .HU r. 2.. .. . .1 CHAPTER IV METHODOLOGY The project which provided the data for this study was designed to investigate adolescent girls' Opinions concerning dress, social ac- ceptance, and related factors.1 The researchers used a number of controls, but recognized the limitations inherent in the study design. Yet, limitations in some areas have led to strength in other areas. This chapter will include the recognized limitations and strengths in discussing: l) the selected community, 2) the selected population, 3) the major variables, 4) the instrument development, 5) the in- strument administration, 6) the methods used to operationalize var- iables, and 7) the methods of analysis. Selection of Community The study chose a Midwestern suburb of 30,198 residents, includ- ing college students.2 The presence of a large university affects the level of education, age, geographic mobility, types of occupations, and income of the population, creating an atypical study setting. 1Joanne Bubolz Eicher and Eleanor Kelley," Adolescent Girls' Viewpoints From Ninth Through Twelfth Grade Concerning Dress, Social Acceptance, and Related Factors," Michigan Experiment Station project #743, research in progress. 2The following community description was drawn from: Betty Wass, "Clothing as Related to Role Behavior of Ninth Grade Girls" (Unpublished Master's thesis, Michigan State University, 1962). 42 43 According to the 1960 census information, community residents who were twenty-five years old, or older, had attained a median educational level of 15.8 years compared to 10.8 for the state. The median population age for the community was 22.2 years, whereas the median state population age was 28.3 years. Geographic mobility was high for the population; 50.7 percent moved into their present homes after 1958. The population atypicality is especially notable in oc— cupations. Seventy-one and nine-tenths percent of the employed residents were white collar, whereas only seven percent were blue collar. As one would expect in a predominantly white collar com- munity, only 12.2 percent had incomes under $3,000, while 31.7 percent had incomes over $10,000. The median community income was $7,152, compared to the $6,256 state median. Although the chosen community is atypical of the state's com- munities, it represents an emerging national trend. The 1960 national census trends included an increased level of education, a higher pro- portion of white collar than blue collar employees, more young adults, and increased residential mobility. Therefore, a white collar suburb whose residents are highly educated and mobile may be a forerunner of United States community trends. The suburb has only one high school, a co-educational public school, which brings students from all areas of the community together in one situation. The incorporation of a low income working class area in the school district two years prior to the project initiation created social class extremes. This served as a major criterion for selecting the project's setting because it offered a situation for a comparative study of the interaction, or lack l l T. L. ‘a § U11 ‘BE lLae: .erac IKE! I y at enal c of in LC‘E ti o . .. .. ... .. I T e a. ... ... a .c s . . Z n r; a: ‘_.a ... 1. 44 of interaction among social class extremes. In other words, it brought together residents of two types of suburbs based upon positive and negative honor in terms of the current life style norms.l:2 An addi- tional criterion which has been valuable during the longitudinal con- tinuation is easy access over time, an important consideration for several reasons. Accessibility has enabled a control over the time of instrument administration each year. It also has given the re— searchers access to the larger community in order to make the ratings comprising the chosen social class index. lResearch reported after the initiation of this study seems to strengthen the importance of the current research setting. Drawing together information from a number of studies conducted in large cit- ies, Farley summarized the trends concerning demographic dispersion in suburban areas. He noted that suburbs retained the general sociologi- cal characteristics which they had in early developmental stages as far back as the 1920's. In other words, upper middle class suburbs re- mained predominantly populated with well-educated white collar residents, whereas lower social class suburbs retained their residential characteristics as less-educated blue collar communities. Contrary to popular opinion, all suburbs do not exhibit upper middle class char- acteristics. Because the current study population contains social class ex- tremes, it offered a situation for comparative analysis of friendship interactions in a setting which draws together residents from suburbs at a polarity in life styles. Moreover, the study taps an emergent dimension cited by Farley, the proportionally greater increase in pop- ulation of higher socio-econmic status. A large proportion of the respondents were not lower class neighborhood residents. The Farley article was drawn from the following source: Reynolds Farley, "Suburban Persistence," in Kimball Young and Raymond Mack, eds. Principles of Sociology (3rd edition) (New York: American Book Co., 1965), pp. 228-239. 2The idea of positive and negative honor as related to style of life was gleaned from: Gregory Stone and William Form, "Instabilities in Status: The Problem of Hierarchy in the Community Study of Status Arrangements," Kimball Young and Raymond Mack, eds., Principles of Sociology (3rd edition) (New York: American Book Co., 1965), p. 113. the researc Beer. .655 ‘ :‘Arv'pv-i t" use; V . tin—— E1": are 4 45 The school situation which offered small population N's enabled the researchers to impose certain additional controls which might have been less manageable in a different situation: using the same re— searchers in administration and data analysis, controlled age, grade, community, school, and sex. Age and grade do change each year; but they are held constant within each data collection period. Population Selection The respondents included all girls in one high school class of 1965. The population was first studied as entering ninth graders in 1962, and restudied each year when the girls were tenth, eleventh, and twelfth graders. The population N's were 154 girls in 1962, 155 in 1963, 143 in 1964, and 138 in 1965. Considering loss and replacement each year, 196 girls were included, and 105 of the original 1962 popu- lation remained in 1965. Only girls were included.1 The study p0pu- lation represents an entire segment of the larger school population which ranged from 1,103 in 1962 to 1,271 in 1965. 1Although Coleman had data for both sexes he separated the boys and girls for sociometric analysis, noting the complexity of trying to integrate the two groups, and stating "the techniques for malyzing such data are in their infancy and until they are better developed, large-scale analysis will continue to have serious omis- sions." Coleman, op. cit., p. 184. e ‘5. £43.85 1112 c D. VL ‘5 ialre i o ‘ A“ KQ‘I. p. 46 Variables The background variables from the larger project in progress, social class, social acceptance and time, are of present concern.l Social acceptance is dependent,and time the independent variable when structural stability or change is the major consideration. Social acceptance is an independent variable when the main focus is social class interrelationships. Instrument Development Responses to the first part of a two part questionnaire developed 2 furnished the currently used data. The question- for an earlier phase naire incorporated questions designed to elicit both social class information and the near—sociometric information. The sociometric information is near-sociometric because the question did not meet all the criteria for a true sociometric question.3 It was developed around the idea of a best friend, defined as "one with whom we spend time and share our secrets."4 This coincides with Jennings' sugges- tion that a sociometric question must have reality for the tested group.5 It also is within Parson's conceptual scheme relating action 1Eicher and Kelley, 0p. cit. 2Wass, op. cit. 33u re, p. 8 for a discussion of the requisites of a true sociometric question. 4See instrument, Appendix B. SHelen Jennings, Leadershipflin Isolation (New York: Longmans, Green Co., 1950), p. 28. ‘ I q . . ”.‘.. ‘ . t.-51ve-: tn especially 1 l I f 1;- 10) C) ”h . n r) O o (v- I m n) 1 C) I 'i n- 11 {I4 “TE h I Q 5““), UQCES a5} .LM'._ ' ‘* - “th04 Seralp H 6V5: 47 to roles,l ipg., the girls were asked to make choices based on one important adolescent role, a best friend.2 The question allowed unlimited choices in the respondent's grade, within other grades in the school, and in other schools, creating a somewhat broader base than many sociometric tests which limit the sociometric response ex- clusively to the interviewed population.3 The extension was deemed eSpecially relevant in a highly mobile population. In addition to the near-sociometric questions, the question- naire included questions about the respondent's family and selected personal characteristics. Several questions were used to develop family social class ratings for each individual. The remaining ques- tions have been used in earlier project phases and will be discussed only in summary fashion as they relate to current findings. Additional 1Parsons, Bales, and Shils. See especially the discussion of role differentiation and goal specification; op. cit., pp. 250-254. Highly relevant is the statement that in the early stages of group interaction there is a process of "jockeying for position" and other- wise settling questions of role status. Ibid., p.250. 2A statement made in discussing dyads is: "for many dyads there are extensive cultural definitions of the roles to be played by each member." Other dyads have few specifications. For example, "expected norms of conduct are fewer and more informal among friends," while the rights and privileges of spouses are regulated extensively by customs and laws. Becker and Useem, op. cit., p. 17. Yet, although there are fewer explicit norms for the friendship role, this study assumed that, among the girls, there are recognizable norms governing the friendship role and regulating behavior in the role. 3When a consideration of the isolates is desired, unlimited choices afford a more accurate picture of group structure. Choice limitation distorts a true picture of the isolates. Lindzey and Borgotta, op. cit., p. 408. data from the questionnair. The t of a two par cafeteria do the question that no one Pledge, eaci. ”Ported an; One I”. for the init the qUEStio: from the 31: assigned fr; Mites. Fr: Slip strUCtv‘ The S 48 data from the school files were used to cross check and supplement the questionnaire. Instrument Administration The questionnaire was administered initially as the first part of a two part questionnaire to the ninth grade girls in the high school cafeteria during an assembly period in February, 1962.1»2 Because the questionnaire sought personal information, the girls were told that no one would be allowed to see their answers. To fulfill this pledge, each respondent was assigned a number and the findings are reported anonymously. One hundred and forty of the 154 ninth grade girls were present for the initial instrument administration. The absentees were given the questionnaire at a later date. The questionnaires were collected from the girls as they were sitting in the cafeteria, and numbers were assigned from 001-154, according to the order of the stacked question- naires. From the close proximity of numbers in the reciprocal friend- ship structures, apparently many girls sat in their friendship groups. The same questionnaire was readministered in February, 1963, 1964, and 1965 when the girls were tenth, eleventh, and twelfth graders; such administration created a similar pattern of student loss and re- placement each year, lpg., at the beginning of the second school term. It was administered in their homerooms by twelve researchers, rather than in the 1962 cafeteria situation. Each year the absentees were 1Wass, op. cit. 2See questionnaire, Appendix B. gmenthe qu= tion proved t placezent we pardon of t allowed a se- sic‘eration x.- the girls. 352 instru: 36:5! Eleve: .CSS and re: 49 given the questionnaire at a later date. This administrative situa- tion proved to be more satisfactory than the 1962 situation. A few modifications in question ordering were made between 1962 and 1963. Because the sociometric question was the key question, it was placed near the first rather than last as it was in 1962. This placement was not crucial in 1962 when the questionnaire was the first portion of the entire battery of questions for which the girls were allowed a seventy minute period; but it was subsequently a major con- sideration when the girls had only a single 15 minute homeroom period to complete the three page questionnaire. Some of the information on the final page was easily obtainable from school files if omitted by the girls. Respondents who joined the p0pulation after the initial 1962 instrument administration were assigned numbers as follows: The 200 numbers indicate girls entering in the tenth grade; the 300 num- bers, eleventh grade; and the 400 numbers, twelfth grade. Considering loss and replacement, a total of 196 girls were respondents during the four year period. Operational Definitions Some data have been analyzed yearly, but this study represents the first longitudinal analysis. The variables were operationally de— fined as follows: The first variable, social class, was operationally defined by assigning family social class ratings to each girl according to Warner's Index of Stat for small cit his social cl be advisable made. First, house type, a dex originall ably would ne‘- also was made Each i dices which \- a to1511 soci 50 Index of Status Characteristics,1 which is considered appropriate for small cities.2 Warner initially used four indices to obtain his social class rankings, but he indicated that modifications may be advisable in some communities.3 Two types of modifications were made. First, only three indices were used: occupation of father, house type, and dwelling area. Source of income, the fourth in- dex originally used by warner, was omitted because the girls prob- ably would not have known this information.4 Second, modification also was made in rating certain educational occupations due to the atypical community educational structure.5 Each individual's family was ranked on each of the three in- dices which were then individually weighted and totalled, giving a total social class rank for each individual respondent's family. This rank included a re-ranking from a previous year when house or 1W. Lloyd Warner, Marcia Meeker, and Kenneth Eells, Social Class in America (Chicago: Science Research Associates, Inc., 1949), pp. 140-1510 2Martindale refers to the Warner method as applicable to small cities. Don MArtindale, American Sociepy (Princeton, New Jersey: D. Van Nostrand Company, Inc., 1960), p. 397. 3Warner, Meeker and Eells, 0p. cit., p. 158. 4The Warner method of determining income is very complex, including factors such as inherited or earned wealth, profits and fees, and salary or wages. .Ibid., p. 142. Income and education were used in the original I.S.C., but in a revised I.S.C. they proved unnecessary. .Ibid., p. 154. 5The modifications were first deve10ped and used by Arlene Bjorngaard in "The Relationship of Social Class and Social Accept- ance to Clothing and Appearance of a Selected Group of Ninth Grade Girls" (Unpublished Masters thesis, Michigan State Univer- sity, 1962). father's 0cm; as well as re: 3636 2838 l f‘. A “:-\-.'D‘. V _ ...... 5’! ‘OSC B 51 father's occupation changed between the administration of instruments, as well as ranking new population additions each year. The individual indices for obtaining a total social class score were ranked as follows: Each girl was asked to list and describe her father's occupa- tion. In addition to her questionnaire information, cross-checks were made in school files and in the city directory. When uncer- tainity arose concerning an individual occupation, resource peOple were consulted to obtain a reasonable ranking according to other similar community occupations. The occupational index was based on a seven point scale with "1" indicating high rating and "7" indicat- ing low rating.1 The residential areas were ranked on a seven point scale from a high of "l" to a low of "7". The seven point ratings for dwelling areas were established around Warner's suggested criteria by a real estate agent, an economist, the project leader (a sociolog- ist), and a member of the initial research team.2 All ratings have been based on the initial rankings assigned to the dwelling areas in 1962. The respondents' houses also were rated by two or more in- dividuals each year. The researchers established the initial criteria for judging the houses in 1962 using Warner's suggested evaluative criteria.3 Each subsequent year a check was made of a 11bid., p. 140-141. 21bid., p. 153-154. 31b1d., p. 149-150. sass-section reorient the to a low of " The cc {aimed by we; was multipli: :izes four. musical so. been made by satin 1963. Social anear-socio: sociometric j a sociogram 1 pairsaand lS( HS recognim Ufa Standarl matrolled d1 \ 115i /. /'. LA) / + . Lind 52 cross-sectional representation from the previous year's ratings to reorient the researchers to each ranking, scaled from a high of "l" to a low of "7". The composite social claSs score for each respondent was ob- tained by weighting the three individual indicesscores. Occupation was multiplied times five, dwelling area times three, and house type times four.1 The computed sum was the total family 1.8.0.2 A low numerical score indicates high social class rank. All ratings have . been made by the same researchers with the exception of one replace- ment in 1965. Social acceptance, the second variable, was established by a near-sociometric question for each individual each year. The near- sociometric information was used to draw a matrix chart and construct a sociogram for each year. Reciprocal friendship structures, mutual pairs,and isolates were determined for each sociogram.3 The research- ers recognize that a criticism of the sociograms has been the lack of a standard method for sociogram construction.4 This problem was controlled during the current study: with the exception of replacing 11bid., p. 124. 21bid., p. 185. 3See definition of terms, supra, p.27-29. 4Lindzey and Borgatta, 0p. cit., p. 411. 53 one researcher in 1965, all matrix charts and sociograms have been developed by the same team according to the same method selected for the initial project phase. In deve10ping the matrix chart, an N by N table was constructed and each girl's choices were recorded as an X. When a choice made by one girl was reciprocated by another girl, each X was circled and a line drawn to connect the two. After all choices had been connected, each group structure was "read" off the matrix by tracing the lines indicating reciprocated ties until there were no more connections. These structures were diagrammed on sociograms, such as the example of the 1962 ninth grade configuration, RFS 107 shown in Figure 1. A circle indicates each girl and lines joining circles indicate reciprocated choices. Each girl's assigned number was recorded in her circle. This method governed the construction of each sociogram. The RFSs were given numbers arbitrarily in the ninth grade but in following years, the numbers were assigned according to the individuals remaining in the RFS from the previous year. The first number indicates year and the other two numbers indicate individual structure. For example, as shown in Figure 2, 107, 207, 307, and 407 indicate the continuity of 107 from ninth through twelfth grades. Some structurassplit between years; therefore, they were as- signed numbers to denote this split, with the larger remaining seg- ment carrying the initial number and the smaller segment assigned a number which indicated its derivation. For example, Figure 2 again shows that RFS 407 and RFS 427 were derived from RFS 107. Because it split, the number 27 was chosen for the less-dominant segment mwma mo mmmHO uoflcom onu mEmuom SOHaz mmmHU Hoozom swam onO as mHuHo ocmuo nunflz onu mcoE< uo3ouo assumed: ozu .h mmm .H shaman eo>_ootua 0.6.3004 <2 toga: o_uu_2 tote; 3 3 OZMOMJ 1?» 0 1m 1 1 1‘ 13112 1a 1E1 131 131 1E1 1E1 131 13' ' 131 18 moma mo mmmHU uoflcow Gnu mfimumm Loflzz mmmHU Hoocom nod: 0:0 as mfluflo mLu mCOE< OonHm>wQ meQEOU h mmm any mm mcoflumcflneoomm mam muHHQm mcfl>ao>m mnu we >HmEESm use» usom .N wusmflm NQN hkk NNN mkk 1\1 B @ 7 kn“. “Kt K‘Hhkk .@ hbmx .. *NHKsQQQ k 191 191 161 33.2... 29531 .0 sons: 22:! =qu ®©0 0.0.0 .290» Olucwd ’ 0 ab 0 moo @ @ ‘b%\ ..b‘hat‘QQ K ’ 1' a \‘We a. a .... a a. O a O 1 a. 1...... m. 1 1 O 1 5.1.0.1 KOx hkt ’ ’ ’ b’ Mu 1 1 9 aka n16 1‘ flu @ .32 .. x3138 u . G 56 since it did not correspond to a number used in a previous year but still contained the number 7. The number 27 indicates that this split of 107 occurred in the tenth grade (1963). Because RFS 107 split many times, the middle number does not indicate year of split consistently for RFS 107, but it is consistently an indicator for the other RFSh. An RFS ending in 0 indicates an entirely new structural formation. For example, RFS 450 was formed anew in the twelfth grade.1 RFSb 330 and 340 were formed in the eleventh grade and therefore constitute a second deviation from the pattern of middle number indicating year of formation. The RFSs with ”teen" numbers in the ninth grade (111-117) either remained intact or com- pletely disappeared; thus they did not present an additional numbering problem. In other words, there is no confusion that RFS 427 could be derived from 117 or 107. This portion of the data was hand coded and tabulated on work- ing c0pies of the sociograms. Lindzey and Borgatta noted that if the N is small, more time may be Spent in preparing for machine analysis than other methods might require to complete the analysis.2 There- fore, hand coding and tabulation were chosen as the most expedient means to accomplish the task. 1The yearly sociograms are presented in the population des- cription, Chapter V, Infra, pp. 73-76. 2Lindzey and Borgatta, 02. cit., p. 419. 57 Usable statistical techniques were limited by the group's nature and size. Because all girls present in each grade were stud- ied, thus constituting entire populations, statistical techniques were limited to descriptive measures. Because the individual RFSE varied in size from 3 to 34, the data cannot satisfy the assumptions of a normal distribution or random sample. In fact, sociometry is a high1y_purposive sampfing technique, and only a statistic which does not require these assumptions, such as a rank order non-para- metric correlation, "is usable. Several statistical sources were explored to determine the factors governing the current use of statistics. Hays noted that "In some practical applications the methods developed using normal theory work quite well even when this assumption is not met, despite the fact that the problem can be given a formal solution only when a normal p0pu1ation distribution is assumed."1 Blair explained why the assumption of a normal curve often cannot be met in social data: The major difficulty in applying the analysis of the normal curve to social data is that in this field the three basic logical principles on which the normal curve is based are not entirely true. A coin or die when freely thrown has an equal chance to fall either way, but in human and social relationships one is not fully free either in heredity or environment.--Certain biases, traits, characteristics, "bents", prejudices, handicaps, and peculiarities tend to persist and to determine the lives of many succeeding gen- erations. 1William L. Hays, Statistics for Psychologists (New York: Holt, Rinehart and Winston, 1963), p. 226. 335 EX Grit c 58 These traits and biases tend to crystallize into customs, traditions, and institutions which are imposed on children from birth and mold and color their thinking, characters, and lives in spite of even strong individual tendencies to "fall" some other way.1 Although inferential statistics are used when the assumption of a normal distribution is not met, the current RFS distribution (3-34) was extremely skewed. For this reason, the researchers decided to omit correlations for which the data could not meet the assumptions. The present study was designed to measure interaction pat- terns of adolescent friendships; therefore, it constituted a highly purposive sampling technique. Blair defined a purposive sampling technique as "one selected to measure some particular phase of a population. A definite purpose or result is in mind. To attain this result, specific controls are required."2 He continued, "Of course, within the limits of the strata and purposes, the sampling is still to be taken at random and without bias."3 Because all possible sam- ples of peer friendship groups are included, his last statement does not influence the current data analysis, i.e., bias of the nature Blair described is not present. lWilliam Myers Blair, Elementary Statistics (New York: Henry Holt Co., 1952), pp. 364-365. 2Ibid., p. 370. 31bid., p. 371. CU 1a.» .'a3 fill]. I1! .U «I; to: 9 & tot 59 Chi-square, as a basis for the contingency coefficient of correlation, is often included as a non-parametric statistic. It does not, however, have the distribution free character, having in- stead a unique family of distributions. Two special Chi-square considerations investigated for this study merit attention. First, expected theoretical frequencies could be established from the 1962 p0pu1ation characteristics, since all hypotheses pose change from the original p0pu1ation characteristics. Thus, when tested in the null, the hypothesis would be no change, or same as in 1962. This method for obtaining expected theoretical frequencies stems from a discussion by wart in which he stated that study hypotheses offer one source of available information usable to establish theoretical frequencies.1 Yet this method asks a different question. It asks how the same population changes over repeated measures. Due to the open character of the social system in the current study, a single population is not present. Rather, there are four p0pu1a- tions which are neither totally independent nor totally related. The second current Chi-square consideration stems from a cautionary note by Hays concerning the pooling of categories. ”This has been done routinely for many years, and many statistics texts advise this as a way out of the problem. [i.e., cells with small 11b1d., p. 371. A .49 1V ..\w 3» . . VZV . 0355 aza. fi s»a: 50:10 .1 ~l1'ilit \ES:-' 5 60 theoretical expected frequencies]. When one starts pooling cate- gories into which observations may fall after the data are seen he is doing something to the randomness of the sample, with unknown . l consequences for his inferences." In the current study unpoOled categories would be difficult to obtain because some RFSb contain only three members, whereas the size ranges up to 34. Thus, the small samples and overlapping p0pu1ation membership led to the re- jection of Chi-square as usable with the current data as a basis for contingency coefficient correlations. In fact, most of the analysis relied entirely on ratios and percentages as the best descriptive statistics available. Reliability and Validity as Related to Sociometry The reliability and validity of the sociometric technique are accepted on the basis of previous researchers' statements. Pepinsky stated that reliability and validity do not mean the same thing in sociometry as in other social science techniques. In fact,she con- tinued, traditional conceptual usage seems to have little direct meaning or application to sociometry. Yet this does not mean that a test-retest measurement is not valuable, only that the purposes for 1J. E. were, C. 0. Neidt, and J. S. Ahmann, Statistical Methods in Educational and Psychological Research (New York: Appleton-Century- Crofts, Inc., 1954), pp. 146-147. Bracket insert is this writer's addition. I aaqu a 0:1“ «f: V vks'u "1: .1. 1,; 31.. C.) ) 3 '§\. 61 which the retest is designed should be defined clearly. Statistical devices should be chosen carefully and factors such as the number of criteria involved in the choices, the number of choices permitted, the number and types of retest situations, and the type sociometric approach should be considered carefully in setting up the methodology for analyzing choice stability.1 While reliability and validity questions may differ somewhat in sociometry, several authors cited studies which indicated that, although individual choices may vary, the patterns of group inter- action remained relatively stable.2 Investigation revealed various analysis factors which seem to make the references questionable as suitable to establish the instruments' usability in a "natural" setting. To illustrate: many studies were conducted with a relatively ”captive" group and the findings were used to restructure the group. Their findings stated that choices remained stable although restructuring occurred. Could the restructuring have introduced an intervening variable which somewhat invalidates the findings for application to a "normal" population? Loomis and Pepinsky raised an additional ques- tion when they stated, ”It is regrettable that none of these writers 1Pauline N. Pepinsky, "The Meaning of Validity and Reliability As Applied to Sociometric Tests,” Journal of Educational and Pay- chological Measurements, IX (1949), p. 45. 2Selltiz, et. a1.,and Goode and Hatt cite the Jennings study, Leadership and Isolation, which was conducted in a girls' training school. Claire Selltiz, Marie Jahoda, M. Deutsch, and S. W. Cook, .Bgsearch Methods in Social Relations (New York: Henry Holt and Com- pany, Inc., 1960), p. 269; and William Goode and Paul Hatt, Methods i§,Socia1 Research (New York: MbGraw-Hill Book Company, 1952), p. 255. has 'L CC! Ahw 1 Q 62 has explicitly stated the extent to which assumptions underlying the use of correlation techniques have been satisfied."1 Jennings introduced a further cautionary note. She stated that high, long-term, test-retest correlations could indicate that "the test was invalid in that it had not caught the flux of psy- chological relations between individuals which are ever in the process of development."2 Lindzey and Borgatta stated that "the user of the sociometric measures seems bound to accept the probability of change over time in the variables he is interested in studying."3 Thus, the issue be- comes more a question of test sensitivity than reliability; because, if one accepts the assumption that human groups are constantly under— going change, only an insensitive test could give a consistent measure- ment from widely spaced time intervals.4 They considered three issues necessary in examining sociometric technique validity: "1. The possibility of limiting one's measurement interest to verbal, interpersonal choice, in which case no validity demonstration is necessary. 2. The possibility of specifying a large number of var- iables which it is intended the test shall measure, in which case the job of relating sociometric response to independent measures becomes relatively unending. 1Charles P. Loomis and Harold B. Pepinsky, "Sociometry, 1937- 1947: Theory and Methods," Sociometry, XI (August, 1948), p. 278. 2Jennings, op. cit. 3Lindzey and Borgatta, 0p. cit., p. 421. 41b1d. 1 . . . . .11. .. . I .... . a . . ~ ... ...—a .e a .. a ..rx. am. end ...—1‘ .P5 911M .. 1 511V ..nu . l . r 111; S O .9 . AU 63 3. The difficulty of interpreting failure of sociometric measures to show a high relationship to independent measures of the variable of interest."1 Northway discussed the relationship of three common measures of reliability and validity to sociometry: "l. The more usual psychological tests assume that they are measuring a trait which remains constant and that if the scores vary this reflects inadequacies of the test rather than changes in the characteristics. . . . Sociometry is based on the assumption that social pref- erenceschange. . . . 2. Another measure of reliability on usual tests is ob- tained by intercorrelating the scores on different test items. . . . Sociometry, however, does not assume apriori that a person chosen on one criterion will be chosen on another. . 3. An individual's score on sociometric tests could be the same on two tests and yet his relationships (in the sociometric structure] might have changed com- pletely."2 The conclusion may be drawn from the above discussion that the refiliability and validity of sociometric techniques cannot be considered it! a.traditional fashion, but must be approached from a perSpective suitable to the assumptions underlying sociometry and the purposes 01? £1 particular study design.3 Since this study primarily focuses on 11bid. 2Mary L. Northway, A Primer of Sociometry (Toronto: University of? Ikaronto Press, 1952), pp. 16-17. Bracket insert is this writer's addi tion. 3Face validity, one traditional method for considering the Yalidity question, can be applied to sociometry, 2.,the assumption is made that the girls did name their friends. The concept of face validity is discussed in Selltiz, et. a1.,0p. cit., p. 165. a: Fb h w 5;; 1 ~ on 64 change in interpersonal relationships, the fact that change in choice status is an expected outgrowth of repeated sociometric measures merits consideration in choosing an instrument suited to its purpose. The purpose is to study pp! group change takes place, as well as whether change occurs. Reciprocity of change in the internal and external aspects of a system are recognized. And little control of factors external to the current study situation was attempted. Parental social class has been considered ,and researchers recognize it as influential in adolescent peer group friendship choices.1 The larger project con- sidered variables such as ecological correlates, academic achieve- ment, popularity and dress, and affiliation of the girls' fathers with either town or university occupations, as influencing change in the internal school situation; however, they are not included in this phase. They are recognized, however, as variables which may influence interaction patterns and will be cited where applicable to the discussion. To reiterate, patterns of interaction and membership change are the main study focus. Summary The project which furnished the data for this study was de- signed.to investigate adolescent girls' Opinions concerning dress, SOCial acceptance, and related factors. The variables considered 1.nthis portion are social class, social acceptance, and time, tile’ Iaackground variables from the larger project. \ 1Supra, p. 7. 65 Each year, from the 1962 ninth grade through the 1965 twelfth grade, a questionnaire designed to obtain social class and socio- metric information was administered. The near-sociometric question was deve10ped around the idea of a best friend. It allowed for un- limited choices in the respondent's grade, within other grades, and other school choices, which gives a somewhat broader base than the many sociometric tests limiting sociometric response exclusively to the intereview population. The study setting was a high school in an atypical Midwestern suburb. Respondents included all girls in the high school class of 1965. The population was studied initially when the respondents were entering ninth graders in 1962. Including the restudies each year, the overall N was 196, with yearly p0pu1ation N's of 154 girls in the l962ninth grade, 155 in the 1963 tenth grade, 143 in the 1964 eleventh grade, and 138 in the 1965 twelfth grade. The following methods Operationally defined the variables: Social class ratings were determined according to warner's Index of Status Characteristics. Three indices were used: occupation of father, house type, and dwelling area. Source of income, the fourth index originally used by Warner, was excluded because it was chaubtful that all girls would know this information. The seven Poillt ratings for house type and dwelling area were established using Warner's suggested evaluative criteria. Sociometric position was established through a near—sociometric question included in the questionnaire. A matrix chart was drawn from each data collection a 0 nd a soc1ogram constructed. 1 1 . <1 . 1 ... _ . . .. . t .11. . h .1, a e .1 A e my. C. .2 . a 01: .1 .t . .1 1. 1 .3 ..1 11 '1. a” .Au .11 .1 A ... » e v.1 «G. Wu. aid in 1 a s . n 1 . .1 o 4 e L» .1 ... :- C. .1. .1» .3 .e .11. a as is .F 114 M1 a... :4 .11. a» . .. a» . a .1. s. 66 Usable statistical techniques were limited by the nature and size of the groups. Since all girls in each grade were studied, statistical techniques were limited to descriptive measures includ- ing correlation techniques which are usable with a p0pu1ation. Be- cause the purposive sampling technique of sociometry yielded diverse sized RFSh, the assumptions underlying certain correlation statistics cannot be met; therefore, the statistical strength was limited pre- dominantly to ratios and percentages. Reliability and validity of the sociometric technique are accepted on the basis of previous researchers' statements. The fact that change in choice status is an expected outgrowth of repeated suaciometric measures is a pertinent consideration in choosing an apprOpriate instrument. Patterns of interaction and membership change irl peer group interaction are the main study focus, i e , it in- vestigated how change takes place as well as whether change takes Place in the peer group friendships. This chapter has discussed the methodology incorporated in stnactying change in the adolescent girls' peer groups. The follow- ing chapters will discuss the findings. The question of whether C:hange takes place is basic to the research design; therefore this q"Jestion also will be answered during the discussion of the findings ConClerning the hypotheses deve10ped to check 31%, L2,, at what 1‘3"€ils of group complexity, change does occur. Obviously if change has not occurred, then the question of M change occurred has no uneaélrting. The next chapter will describe the four year p0pu1ations, fol- 1. ()‘Vebd.by chapters reporting whether, where, and h9W change OCCUTYEd- CHAPTER V LONGITUDINAL POPULATION DESCRIPTION The longitudinal p0pu1ation composition shifted over the four years from three perspectives, p0pulation size, social class com- position, and social acceptance structural composition. This chapter will describe the p0pu1ation from the three cited perspectives. POpulation Size The p0pulation changes in size are a result of out-migrationl of? some members and in-migration of others, not an unexpected finding it! an open social system. A complete breakdown of the p0pu1ation changes in size from 1962 through 1965 is presented in Table I. There were 154 girls in In 1963 there were 155 tenth grade girls. How- the 1962 ninth grade. ever,. 20 ninth graders migrated from the social system and 21 new girls migrated into the social system. 1Because the subsequent discussion will include a tracing of :0Vement between the sub-structures, or RFS's, within the social system 8 Well as movement from and into the system per se, the term migra- Vvill be used to denote residence change which either places the tiOn git]; in the social system (in-migration) or removes her from the C161]. system (out-migration), whereas the term shifting will be used tc: enote within-system movement from one RFS or RFC to another. 67 . ..11. 1. 1 1-_ ‘o ag- \ 68 a Table 1. Four year numerical summary of the stability and change in the p0pu1ation of girls in one high school senior c1ass,1965 1962 1963 1964 1965 Stability and Change Ninth Tenth Eleventh Twelfth Grade Grade Grade Grade N N N N Stability: Longitudinal stability (N reSpondents remaining from 1962) 13411:b 116* 108* Between year stability (N respondents added after 1962 who remained in population) 21* 14* 22* Total stability per year 155 120 127 Channge: DJ out-migrants from 1962 20 19 10 N new migrants into each year's jpopulation (after 1962) 21 13* 8* N out-migrants from previous year -- 7 5 Total yearly change (loss and re- placement) 41 39 23 \ T°tal N's 154 155b 143b 138b \ a. aThe summary tables in the text are indicated by Roman numerals nd r~the detailed tables in Appendix B by Arabic numerals. allc: bTotal N derived by adding * numbers each year. The discrep- gr ies in the eleventh and twelfth grade columns are due to ninth tuide respondents who migrated out in a between year and later re- 19 lied to the p0pu1ation, $5., in the eleventh grade, 116 remaining + thelost = 135, not 134 as stated in the tenth grade column. Also, in terlfth grade, 108 + 10 = 118, not 116 as presented in the eleventh Eglréié162 column. 69 In 1964, the eleventh grade population decreased to 143.girls, twelve less than the 1963 tenth grade group. From the original 154 1962 ninth grade social system, 116 members remained. Twenty-six tenth grade p0pu1ation members were no longer present in the eleventh grade. Of these, seven were new girls in the tenth grade and thus were respondents only one year. Thirteen eleventh grade respondents migrated into the social system as eleventh graders. By 1965 the twelfth grade population decreased to 138. One hundred and eight remained from the initial 1962 ninth grade social This includes three girls who migrated out in a between year system. Fifteen migrated from the eleventh grade popula- and returned later . t:ion.and eight new girls migrated into the social system when they snare 1965 seniors. Five eleventh grade additions migrated out before the: twelfth grade study and thus were in the social system only one Year. In summary, considering loss and replacement, the longitudinal stniciy included 196 reSpondents who were members of the social system SC"n€! of the time between the 1962 ninth grade study and the 1965 twe 1 f th grade study . Longitudinal Social Class Composition As stated in the methodology chapter, the social class ranking ‘Vaas Dnade for each girl each year, including re-ranking when apprOpriate Re"rankings included both changes in father's occupations and changes 1 :1 addresses. However, the predominant number of changes resulted in \ 1Supra, p.50. 70 movement within a social class category rather than markedlydiffer- ent social class scores leading to a change from one social class category to another. Only five social class re-erankings resulted in a change from one social class category to another; all five occurred in the eleventh grade study, and resulted from occupational changes. Three of the five changes involved girls who listed their mothers as the main sources of income. Four resulted in downward mobility from upper to middle social class, whereas the fifth resulted in upward mobility from lower to middle class. Contrary to the between category re- r-ankings, the within category re-rankings included changes in residence as well as changes in occupations. Table II summarizes the social class composition by yearand grade. In the 1962 ninth grade there were 16 lower social class mem- bers, 48 middle social class members ,and 90 upper social class mem- bers, creating a skewed distribution along the social class variable. 33' the 1965 twelfth grade study, the distribution became more distinctly Skewed by the lower social class category. Only five lower social The middle-and upper social classes re- class respondents remained. Thus, the predominant mained fairly stable at 45 and 88 respectively. change from an N of 154 in the ninth grade to an N of 138 in the twelf-th grade is due to lower social class loss, not an unexpected finding in light of school dropout literature. This point is reinforced b y a check of the school files which revealed that five of the eight Ag Av 1H“ t 1- b r 11 .n u «L 31.1» urn. ‘ c a: RAD Lin 1 .1‘ .e . "I «L b u c . n\ ‘- ”J ayg ‘1‘11 10.. 71 1 were lower social class. There were two upper known "true" dropouts social class and one middle social class drOpouts. Social class rankings ranged from 12 to 84, with low score indicating high rank. Appendix B contains yearly tables (Tables 1 - 4) with the complete p0pu1ation distributions by reciprocal friendship categories and individual RFSh. Tables 5 - 8, also in Appendix B, con- tain a complete distribution of the scores by individual indices. Table II. Four year summary by percentage and number of the social class distribution among the girls in one high school senior class, 1965 Year and Grade 1962 1963 1964 1965 Ninth Tenth Eleventh Twelfth Social Class Grade Grade Grade Grade N X N 1 N Z N Z Upper 90 59 94 61 93 66 88 64 Middle 48 31 48 31 43 29 45 32 LOWer 16 10 13 8 7 5 5 4 Total 154 100 155 100 143 100 138 100 Williams found, in analyzing the individual indices for the 1962 ninth grade population, that more isolates and mutual pairs had I‘DVVGEI: ranked dwelling area scores than RFS members. However, gener- ailil}? the dwelling area scores ranked higher than house type scores ._________ 1Defined as one who withdrew from school without entering £1 I1Other school system, or "quit" school. 72 for the population as a whole. When a composite ISC score weighting and totalling the three indices was drawn, there were fewer mutual pair members from upper social class backgrounds than isolates or RFS members.1 Paralleling William's ninth grade findings, Hendricks also found in the 1965 senior class that the dwelling area scores were higher than individual house type scores.2 Thus the girls continued to live in areas superior to their individual residence. However, by the twelfth grade Hendricks found that the isolates ranked lower on all social class individual indices, indicating an apparent relation- ship between social class status and group membership.3 Longitudinal Social Acceptance Structural Composition Group membership, or acceptance, in this study is based upon reciprocated friendship choices. The choices were used to construct sociograms according to the method described in Chapter IV.4 The yearly sociograms, presented in Figures 3 - 6, show that the popula- tion composition shifted in number, size, and shape of configurations. As previously stated, the population N's shifted from 154 in the ninth grade to 138 in the twelfth grade. The yearly N's are distributed 1Madelyn Williams, op. cit., p. 56. 2Suzanne Hendricks, op. cit., p. 30. 31bid. 4Supra, pp. 53-57. RFS [/6 v” xsourtsl a v 1501 AYESZJIGNM ‘ Om G v O @ a! A G G v xsaurtsr SELF ‘ ‘ v /501 4 7:5, — 'cawrusw' a @ 0 ® Q @ @@ @@ GO a M |L n. I- a. mfiffwfiafia 969G; . MPG s m k 3 1E1 151 1.1 Reciprocal Friendships of the Ninth Grade Girls in One High School Class Which Became the Senior Class of 1965 Figure 3. 11?? © RC:2 R e3 i C £1131 181 “1‘81 § 11“”81 1|1 1g1 l s 151 181 1.1 191 @ © 11§ 151 1B1 1B1 O 161 RF5237 000 030 Eocene _ @000 «mono QC} is; sees .00 sue ea —~rRU£~ USELF r551 £53323 LEGEND Reciprocal Friendships of the Tenth Grade Girls in One High School Class Which Became the Senior Class of 1965 Figure 4. NNm. hkk I." 3 NR. WKQ OVM.hkk hag ®O «Hwo’ mno ’hnm. mkfi 131 191 131 lb- 2 “I I @Qggw men wuo M 181 1E1 asses ‘z I mood mo mmmHo meadow onu weapon soda: nmmHo Hoonum Amwm 0:0 as mango uvmuo suao>mam on» no moanmpnoAum Hmoouaaoom .m ousmam ozuouq ...»..an 22:34 thhkk Dnnhkt HHMWKQ @Om. 9®m ®@@ ’ ’ ®@@® , . ’ ’ ’ , @fififia . ‘ ‘ ‘ ‘ . amateu- 1 4 Pt u 6% @@®® ©®®G - ‘ . .3me n nub-Sb ®@@@@ @@®@® .3163: 1 N P: q 3% seem sass . SE. 1 42:89 1s moo." mo 2:30 nowcom docs—Om soda ’5 sun 3qu 009.5 savage on» no wmqnucdowuh 13008303 .0 ounmwh oses©91éa @ o o. sings H see. a :3 a 5.3 a G .O@®®O@ Iv 25.3 32.23.13.3363 8:. .8 a 1 .33.: 2:83 QO .....Q m... a: G a: ....hhx‘qofi .03.: 2'38 .023 8 mos 8‘ h.“ ”9‘40 44.000 nru KNV hkk ./ \1/ O Q mg @ nmxo.) r @ . 0V WKQ ‘59 “KC romxT 0 ® 77 among the reciprocal friendship categories in the following structural patterning presented in Table III. Table III. Four year numerical summary of the p0pu1ation structural composition in the reciprocal friendship categories formed by the girls in one high school senior class, l965 Year and Grade 1962 1963 1964 1965 Structural Composition Ninth Tenth Eleventh Twelfth Grade Grade Grade Grade Number of isolates 23 27 36 20 Number of MP5 9 8 9 8 Nimber of RFS structures 17 13 15 10 'Iotal number of structuresa 26 21 24 18 (Total number of dyads in the (131) (l28) (103) (124) structures)a aMP's and RFS's As shown, the category N's remained fairly stable with the ex- Cer>tion of the eleventh grade when there are more isolates and fewer dyads, indicating fewer ties among the eleventh grade population. The! number of dyads in the senior class was almost the same as the nunflber in the ninth grade when there were more structures and a larger pCHDtllation N. Thus, there is an apparent increased integration in larger, more complex structures by the twelfth grade, a reversal of ttIEI eleventh grade trend to more, smaller groups. A scanning of the year 1y sociograms in the previously introduced figures, Figures 3 - 6, ‘0' . . . . . . . 151]. fac1litate Visual comparison of the shifting compOSitions. A 78 more complete discussion of the population composition along the social. acceptance variable will follow in Chapters VI and VII which trace change and stability of the reciprocated relationships formed by the girls. CHAPTER VI DISCUSSION OF FINDINGS: WHETHER AND WHERE POPULATION FRIENDSHIP CHANGES OCCURRED The hypotheses formulated to investigate p0pu1ation change guided the analysis. The first concern was parental social class ranking as a possible factor influencing peer group formation. Next, three as- pects of adolescent girls' peer group change were studied: structural patterns, membership positions, and category sizes. This chapter, «organized around the five general hypotheses presented in the previous (fliapter, will report the findings obtained concerning whether and where change occurred . Social Class Composition The first series of hypotheses was deve10ped to explore whether mental social class rank will influence an adolescent's sociometric 2081tion. one <3f the major criteria in selecting a population to study. As stated in Chapter IV, social class extremes served as 1 The sub- hypotheses, following stratification literature, predicted that ‘more RFS's would be homogeneous and, of the mixed groups, the differences ‘flhllcl be ”variations around the means" rather than sharply different- lated social class rankings. A five point difference in social class \ 1Supra, p.43 . 79 80 score from high into middle, or low into middle, was considered "var- iation around the mean." The homoegneity of an RFS was determined by raw scores. A scanning of the individual RFS mean scores revealed that some groups' means were pulled up, or down, into homogeneity by the extremely high, or low, scores of other members. To establish homogeneity by mean scores would, therefore, distort a true picture of social class differentiation. The individual RFS's from 1962 through 1965 exhibit homogeneity, "variation around the mean," and distinctly different social class compositions. In the ninth grade, presented in Table IV, there were six homogeneous RFS's, three "variations around the mean,’ and eight with distinctly different social class rankings. But, six of the eight with distinctly different social class compositions were "skewed" by one individual. They include RFS's 107, 109, 110, 111, and 112, each containing five or more members, and 116 which had only three members. The two RFS's containing a wide range of social class scores, RFS's 105 and 115, were small, three or four member groups. The mean score of RFS 105 was low enough to pull the group into the low social class category although two girls had middle social class scores. Table V shows that by the tenth grade the number of RFS groups had decreased from 17 to 13. Considering this decrease in number, essentially the same prOportion of groups as in ninth grade fall in each of the three designated social class composite classifications: four homogeneous, two "variations around the mean," and six distinctly different social class compositions. Three of the groups with 81 {Table IV. Summary by reciprocal friendship structures of the 1962 ninth grade social class composition among the girls in one high school senior class, 1965 RFS Number Social Class Scores Number in Social Class Social Class Homogeneity Range Moan Upper Middle Lower 101 60-70 64 3 -3 102b 26-68 39.6 7 7 1 * 103 12-43 32.5 4 2 9: 104 29-43 35.7 7 1 * 105 45-75 64.0 1 3 + 106 19-35 25.5 4 - 107 161-49. 31.1 17 10 + 108 19-35 27.1 4 - 109 33-47 40.5 6 2 + 110 26-50 42.1 3 2 l + 111 12-48 34.0 2 2 + 112 27.5-51 37.2 2 3 + 113 21-32 27.9 7 - 114 37.5-22 28.5 3 - l 15 19-54 41.0 1 2 + 116 19-41 31.0 2 1 + in 27.5-36.5 31.8 3 - 1"Eat-1 social class score: 43 \ .5 a Key: - Homogeneous social class group. * Social class composition "variation around the mean." + Composed of distinctly different social classes. bNote: RFS's 102, 107, 109, 110, 111, 112 skewed by one person's .markedly different score. Social class homogeniety was eétablished by'individhal'social class scores, not group mean. Variation around the mean was determined by a 5 point range within the middle social class score range at either end of the scale. 82 treble V. Summary by reciprocal friendship structures of the 1963 tenth grade social class composition among the girls in one high school senior class, 1965 __,, li‘ RFS Number Social Class Scores Number in Social Class Social Class Homogeneity Range Mean Upper Middle Lower 12(31 58-72 66.5 1 3 *3 12(32 12-50 34.3 18 16 + 204 16-41. 29.86 3 l * 205 24-75 51.5 4 3 5 + 207b 12-43 27.9 10 2 + 213 12—35 26.9 18 - 216 33-41 35.7 2 1 * 217 25-61 34.5 4 1 + 219 22-33 27.3 3 - 220 32-35 34.3 3 - 222 26-35 19.7 6 - 227 39-66 50.3 2 1 + £37 26-49.5 35.7 3 2 + Mean social class score: 33.05 \ aKey: - Homogeneous social class group. * Social class composition "variation around the mean." + Composed of distinctly different social classes. bNote: RFS's 207, 217, and 227 skewed by one person's markedly different social class score. 83 distinctly different social class composition were skewed by one per- son: RFS's 207, 217, and 227. All three groups have mean scores which fall within the social class range of the dominant membership. Two groups which had more than one member in distinctly different social class rankings, RFS's 237 and 202, also had sufficiently large, or small, scores to pull their mean scores into the dominant group. But RFS 205 continued to exhibit a wide range of social class scores which becomes more pronounced in the tenth grade with increased membership from a four person to a twleve person group. The eleventh grade social class data are summarized in Table VI. The eleventh grade increased number of RFS's with fewer members per group continues to be reflected in the social class data. Seven RFS's were homogeneous, two exhibited "variation around the mean," and six were distinctly different in social class composition. Only one, RFS 332, was skewed by one person. She represents a rank change from upper to middle and probably continued to be perceived as upper by her peer mates. The remaining five groups were widely distributed to the extent that only one RFS, 337, was sufficiently high to pull the mean RFS score into the upper social class range. Thus, more RFS's appeared with diverse composition than in previous years. Perhaps this diversity tends to reflect Gordon's idea of "clique” breakdown due to competition for positions - a different idea from the Parsonian thesis of jockeying for position in the early stage of group formation, unless it is considered as a reformation around changing group inter- ests. The dating age may explain the shifts, or as Gordon stated, competition for honors.1 1§22£29 p. 18 for a discussion of Gordon's findings. - ...—m m_’__—s.° .J-E . x I...» \ 84 Table VI. Summary by reciprocal friendship structures of the 1964 eleventh grade social class composition among the girls in one high school senior class, 1965 RFS Number; Social Class Scores Number in Social Class Social Class Homogeneity Range Mean Upper Middle Lower 350 38-48.5 43.2 3 -a 305 26-70 47 1 l l + ' 342 29 4 - 307 12-41 26.4 13 2 * 313 21-38 29.2 4 1 * 317 27.5-57 37.5 4 2 + 322 19-56 38.2 6 5 + 330 25-34 29.7 3 - 3321) 19-45 31.0 13 4 + 333 26-33 30.3 3 - 335 34-72 49.3 1 2 l + 337 26-49.5 35.7 3 2 + 340 45-55 49 3 - 347 21-35 28 4 - 367 16-33 25.7 3 - Mean social class score: 25.72 aKey: - Homogeneous social class group. * Social class composition "variation around the mean." + Composed of distinctly different social classes. bNote: RFS 332 skewed by one person's markedly different score. She was initially an upper social class person whose rank changed in the eleventh grade. 85 The final social class table, Table VII, contains the distribution of social class ranks for the 1965 twelfth grade RFS's. The distribution shows that only one homogeneous RFS emerged, a derivation from RFS 107, RFS 427. Three RFS's, 414, 452, and 477, contained scores which are merely "variations around the mean." Of the six RFS's with distinctly different social class compositions, four are skewed by one person: RFS's 401, 407, 422,and 460. The deviant score in RFS 422 represents a second example of the few social class changes from one category to another in the four year study. She was initially an upper social class person and probably continued to be perceived as such by her peers. Both of the RFS's with diverse social class scores, RFS's 417 and 432, continued to have a mean score in the upper social class range. The total RFS mean scores for each year show an increase in mean social class ranks1 over the longitudinal study, reflecting the loss of all except four of the initial 17 lower social class members. Five were known "true" school dropouts, that is to say, students who withdrew from school and remained in the community without enter- ing another school. The remainder of the low social class respondents lost from the population moved and their later school status is un- known. It seems reasonable to assume,in view of school dr0pout liter- ature, that additional low social class members became dropouts from the school system into which they migrated. The five low social class dr0pouts represent two-thirds of the total dropouts over the 1Low numerical score indicates high social class rank. 86 Table VII. Sumary by reciprocal friendship structures of the 1965 twelfth grade social class composition in one high school senior class, 1965 RFS Number Social Class Scores Number in Social Class Social Class Homogeneity Range Mean Upper Middle Lower 401 46-76 62.6 2 3 +3 407b 12-49.5 26.2 14 3 + 414 26-37.5 30.2 2 1 * 417 21-57 34.1 18 6 + 422 12-56 31.4 4 1 + 432 12-72 33.1 18 9 1 + 460 26-50 36.0 3 1 + 452 18-43 31.4 3 2 * 477 16-40 25.0 6 1 * 427 38-46 41.5 4 - Mean social class score: 33.26 aKey: - Homogeneous social class group. * Social class composition "variation around the mean." + Composed of distinctly different social classes. bNote: RFS's 401, 407, 422,and 460 are skewed by one person's markedly different score. The single middle class person in RFS 414 is actually on the break with a score of 37.5. The deviant score in RFS 422 represents a social class change. She was initially an upper social class person and therefore is probably still perceived the same by her peers. 87 longitudinal study, not an unexpected finding in a p0pu1ation which has an expectation that approximately 95 to 98 percent of the member- ship will attend college. The mean social class rank for the eleventh grade RFS's is higher than any other year, although there are fewer total popula- tion members in the twelfth grade. The factor of large isolate num- bers and fewer RFS members in the eleventh grade group has been re- peatedly established as influencing all aspects of the findings. A higher RFS mean social class score when there are fewer RFS members could foster a tentative hypothesis that there is a relationship be- tween social class score and group membership. Specifically, the fewer the group members, the higher the incidence of upper social class group members in preportion to the total number of upper social class respondents in the p0pu1ation. A comparison of the four years' social class distributions by RFS's shows that approximately one-third to one-half of the RFS's each year contained diverse social class scores. Yet a more detailed com- parison reveals that the majority are skewed by the lower or higher score of one member only. Therefore, very little interaction between distinctly different social classes existed in the four year longitud- inal study. The assumption seems to follow logically that the data tend to support the stratification literature stating that adolescents form their peer groups along parental social class rankings. The 88 data also tend to support warner's idea that social classes tend to merge into contiguous classes along certain interaction dimensions.1’2 Change in Peer Group Friendship Patterns The second set of hypotheses was developed to determine where change occurred in the patterns of peer group relationships. Specific- ally, the hypothesis stated that the populations' peer group friend- ship patterns will exhibit more change than stability over a four year time Span. Change in the RFS's present longitudinally was hypothesized as resulting from loss of ninth grade structures and creation of new structures in subsequent years. Table VIII traces the continuity and change¢mfthe structures from the ninth through the twelfth grade.3 Of the 17 structures present in the initial ninth grade popu- lation, seven,or 40 percent, remained in the twelfth grade, including several small structures, RFS's 103, 106, and 116, which united in RFS 432. The three structures, plus RFS 114, joined the initial RFS 102 structure to form the large, 34 person, RFS 202 in the tenth 1Supra, p. 38. 2A formal acceptance or rejection of the hypotheses by the data is not possible without the formal testing with inferential statistics. As stated in Chapter IV, the distribution of the popu- lations within the study framework limits statistical usages largely to percentages and ratios. Supra, pp. 58-59 . This factor governs the acceptance or rejection of all subsequently discussed hypotheses. 3Supra, p.53 and 35 for a statement of the rationale used in numbering the structures, and p. 38 for the hypotheses deve10ped to guide the analysis. 89 Table VIII. Four year longitudinal continuity of the reciprocal friend- ship structures among the girls in one high school senior class, 1965. Year and Grade 1962 1963 1964 1965 Ninth Tenth Eleventh Twelfth Grade Grade Grade Grade ca L F c L F c L F 101 201 - - - 201 - 401 - - 102 202b - - - 202 - - - - 103 202 - - 332 - - 432 - - 104 204 - - - 204 - - - - 105 205 - - 305 - - - 305 - 106 202 - - - 106 - — - — 107 207 - - 307 - - 407 - - 108 - 108 - - - - - - - 109 - 109 - - - - - - - 110 - 110 - - ,. - - - - - lll - 111 - - - - - - - 112 - 112 - - - - - - - 113 213 - - 313 - - 417 - - 114 202 - - 332 - - 414 - - 115 202 - - 332 - - 432 - - 116 216 - - - 216 - - - - 117 217 - - 317 - - 417c - - - - 219 - 219 - - - — - - 220 — 220 - - - - - - 222 322 - - 422 - — - - 227 - 227 - 427 - - - - 237 337 - - - 337 - - - - - - 330 - 330 - - - - - - 342 - 342 - - - - - - 340 - 340 - - - - - - 347 - 347 - - - - - - 333 - 333 - - - - - - 335 - 335 - - - - - - 367 - 367 - - - - - - 350 - 350 - - - - - - - - - 452 - - - - - - - - 460 - - - - - - - - 477 8Key: C = continued, L = lost, F = newly formed. bThe numbers are traced by the dominant number into which the small RFS's moved. In the case of RFS 202, there are an equal number from RFS 102 and RFS 103. The two derivations were chosen to eliminate confusion with RFS 113 which continued. RFS 112 was lost in 1963. cRFS 417 is actually a coalition of RFS 117 and RFS 113 members in the 1965 p0pu1ation. 90 grade. In the eleventh grade, most of the initial membership of RFS 102 disintegrated into isolates and mutual pairs, leaving RFS 332 essentially composed of the small structures which joined the group in the tenth grade. One of the small structures, RFS 114, split from the group again in the twelfth grade as a three-person chain with two members from RFS 114 and one member from RFS 106, which also joined RFS 202. In tracing the loss of ninth grade RFS's by years, Table VIII again shows that 5 structures, or 29 percent were lost by the tenth grade, and 10 structures, or 60 percent were lost by the eleventh grade. The continuity of 60 percent loss in the twelfth grade stems from a balancing loss of RFS 302 and regaining of RFS 401, one of the two low social class structures present in the longitudinal study. Fifteen new structures were formed between the tenth and twelfth grades, representing the equivalent of 88 percent of the initial 17 ninth grade RFS's. One factor which explains a large proportion of the newly created groups is the splitting and recombin- ing of some ninth grade structures. Table VIII again shows that many of the subsequently created structures carry numbers indicating derivation from RFS's 102, 107, and the 117-113 union. The Figure 6 sociogram for the twelfth grade reveals that they were the larger 1 Only five entirely new structures, the senior class structures. RFS's ending in 0 numbers, were formed between the tenth and twelfth grades. 1Supra, p.76, 91 When the findings of the above paragraphs are summarized, a picture of structural change and stability emerges as follows. Forty percent of the initial ninth grade structures remained in the twelfth grade, sixty percent of the twelfth grade structures were not present in the ninth grade. Therefore, the hypothesis that there would be more change than stability in the RFS's from the 1962 ninth grade through the 1965 twelfth grade is supported. Statements made by the reapondents seemed to indicate their accurate perception of a highly structured sub-system within their school population. They used the term "clique" frequently. A num- ber of the girls stated that a new person found it difficult to gain acceptance in a "clique" or group because many of their groups had formed in elementary school and carried over into the high school setting. One factor which may explain the acceptance of some "new" girls (as denoted by 200, 300, and 400 numbers in their configurational circles) is the fact that many are returnees and not entirely new to the p0pu1ation. This circumstance stems from the educational structure of the p0pu1ation, in which many fathers earn sabbatical leaves, or are away for one or two years on research, or consultant positions outside the community. Information available through the other school choices made by the respondents seems to indicate, however, a re- turn to structures other than those to which they previously belonged. Perhaps this shift may be explained by changing interests through new experiences away from the school p0pu1ation and shifting interests of their former peer-mates while they were gone. Because only a few 92 girls were added as re-entering rather than new girls each year, no specific cases will be mentioned. An awareness of the ecological correlate as an important factor influencing the acceptance of a new girl was reflected in the state- ment of one respondent. When asked what factors were important for a new girl to get in with the p0pular girls, she replied, "Get in the right car pool." Thus it would seem that movement into the "leading crowd" may be easier for a new girl moving into the "right" neighbor- hood, whose past pattern of living is unknown, than for a girl whose family mobility occurred geographically within the community. After establishing that the RFS's did change between the ninth and twelfth grades, the second set of hypotheses relates to change in size. Following Hallworth's findings, it was hypothesized that the RFS's would become larger in size and therefore fewer RFS's would be present in the twelfth grade than in the ninth grade. The two hy— potheses seem to follow one another logically if they are related to the Parsonian idea that changes in one sub-system must be balanced by change in another.1 Table IX contains a complete distribution of the longitudinal p0pu1ation's RFS's,giving number and size of each configuration from ninth through twelfth grades. Seventeen RFS's appear in the ninth grade and only 10 in the twelfth grade, giving a 41 percent decrease in number. Between year comparison shows a finding similar to Gordon's 2 findings. The pattern reversed between the tenth and eleventh grades, 1Supra, ppm 4-5 for a discussion of Parson's idea. 2Supra, p. 18 fOr a discussion of Gordon's findings. 93 Table IX. Four year numerical summary according to membership size of the reciprocal friendship structures among the girls in one high school senior class, 1965 Year and Grade 1962 1963 1964 1965 Number of Ninth Grade Tenth Grade Eleventh Grade Twelfth Grade Members in Number of Number of Number of Number of Group Such Groups Such Groups Such Groups Such Groups 3 5 4 6 l 4 4 2 3 2 5 l 2 2 3 6 2 l 1 7 l l 8 2 ll 1 12 2 15 1 1 17 l 1 18 l 24 l 27 l 28 1 34 1 Total N of groups l7 13 15 10 Total N of members 13 112 89 102 Mean size: 6.65 9.54 5.93 10.2 Median size: 4 2 3 3 Mbde: 5 4 6 3 94 iafia: numbers increased from 13 to 15 RFS's. The finding supports Gordon not only in an increased number of structures, but also in a decreased number of respondents in structures. The between year reversal of patterns also carried over into mean group sizes, not an unexpected finding when a sub-unit systemic balanced relationship is considered. Obviously, with fewer respondents and more structures, the mean size per group will decrease. A com- parison of the ninth and twelfth grade mean group sizes shows an increase from 6.65 to 10.2, or a difference of 3.35 persons per group.1 Thus, the mean size in the twelfth grade represents an in- crease of 53 percent over the ninth grade size, supporting the hy- potheses formulated from Hallworth's findings. The group sizes in- creased and the groups became fewer in number. This becomes a rather interesting finding when the ages of the two populations are con- sidered. Hallworth's study included more elementary pupils, whereas this study encompassed only high school age students. Perhaps the finding also supports the respondents' accurate perceptions of elemen- tary school groups continuing as peer groups in the current high school situation. If the groups become larger in size and fewer in number, then logically, the configurational patterns will change, as the hypotheses posited. Table X includes not only the RFS configurations but also the mutual pairs and isolates; however, only the RFS's are considered in the hypothesis. Table X lists six general configurational classif- ications, but each complex structure represents a different pattern. 1This is based on absolute means, recognizing that one cannot actually have a third of a person. 95 Table X. Four year numerical summary of the structural patterns formed by the reciprocated friendships among the girls in one high school senior class, 1965 - w Yea£_gnd Grade 1962 1963 1964 1965 Structural Ninth Tenth Eleventh Twelfth Patterns Grade Grade Grade Grade INa SN IN SN IN SN IN SN Isolates 23 - 27 - 36 - 20 - Mutual Pairs 18 9 l6 8 18 9 l6 8 Chains l6 4 23 6 28 8 7 2 Triangles 16 5 - - 7 2 - - Triangle-chain - - 9 2 6 l 12 2 Wheels (including broken wheel) 19 3 - - 5 l 9 2 Square - - 4 l - - 5 1 Complex structures 62 5 76 4 43 3 69 3 Total N: 154 155 89 138 Total Structures: 17 13 15 10 aKey: IN: Individual Number SN: Structure Number Eight patterns were present in the ninth grade, including five complex structures. In the twelfth grade only four of the initial patterns were present, giving a 50 percent difference in the comparative 1962 and 1965 sociometric sub-unit patterns. MOst of the difference is due to the shifted patterns in the complex structures. Two new basic structural patterns appeared in the twelfth grade and only one ninth 96 grade pattern was missing, the three person triangle. This omission reflects the factor of increased mean size per group as the groups de- creased from freshman to senior year. The eleventh grade reversal of pattern is also evident in configurations: three complex structures with fewer respondents in each and more small structures. According to Hallworth, as the structures increase in size and become fewer in number, they also become more cohesive. Hypotheses were posited in this relationship although it directly Opposes Kretch, Crutchfield,and Ballachy's statement that high cohesion and small size correlate. The data supported the latter statement and not the posited hypotheses. Table XI includes a complete longitudinal distribution of all RFS's and segments of RFS's ranked according to cohesion.1 The data in Table XI are combined in Table XII to present yearly mean cohesion scores. The decision was made to check the correlation between size and cohesion first, because if no relationship exists, then logically a comparison between years would be of little value. Spearman rho correlations2 revealed high correlations of .88, .85, and .65 for ninth, tenth, and eleventh grades, respectively. But in the senior year the correlation is only .28. This correlation seems to reflect the increased mean group sizes which leads to lower cohesion. However, 1Supra, p. 32 for the formula used to obtain the cohesion scores. 2See Sidney Siegel, Nonparametric Statistics (New York: MCGraw- Hill, 1956), p. 209, for the formula used. 97 Table XI. Four year summary of cohesion scores distributed by in- dividual reciprocal friendship structures among the girls in one high school senior class, 1965 Year and Grade 1962 1963 1964 1965 Ninth Grade Tenth Grade Eleventh Grade Twelfth Grade RFSa RFS RFS RFS RFS RFS RFS RFS RFS RFS RFS RFS S C S C S C S C 101 3 100 201 4 92 367 3 100 427 4 75 114 3 100 204 4 83 305 3 83 422 5 70 116 3 100 216 3 83 340 3 83 414 3 66 117 3 100 219 3 66 350 3 83 452 5 65 105 4 83 220 3 66 335 4 66 460 4 60 110 6 66 227 3 66 333 3 66 401 5 50 115 3 66 237 5 60 330 3 66 477 7 47 108 4 66 217 5 45 342 4 58 407 l7 18 103 6 60 222 6 43 313 5 55 417 24 13 106 4 58 205 12 27 347 4 50 432 28 12 112 5 55 207 12 20 317 6 47 109 8 55 213 18 17 337 5 40 111 4 50 202 34 9 322 ll 30 113 7 33 332 17 21 104 8 27 307 15 20 102 15 18 107 33 12 b102A 53 205A 43 322A 70 407A 29 1028 32 2053 40 3223 57 407B 50 107A 28 213A 65 332A 66 417A 31 107B 22 213B 47 332B 66 417B 50 213C 43 332C 52 417C 44 202A 24 307A 40 432A 20 202B 36 307B 39 432B 100 2020 27 4320 50 8Key: RFS: Identifying number of RFS. RFS S: Number of members in RFS, i.e., size. RFS C: Cohesion score for RFS, based on a possible total of 100. bSegmental breakdowns of larger RFS's with cohesion scores for the segment. 98 some of the difference originates in more out-system choices, a factor supporting Gordon's finding concerning group weaknesses as the end of high school education approaches.1 Table XII. Four year mean cohesion scores of the individual reciprocal friendship structure cohesion scores among the girls in one high school senior class, 1965. Year and Grade _ 1962 1963 1964 1965 X Cohesion Scores Ninth Tenth Eleventh Twelfth Grade Grade Grade Grade Total RFS's i cohesion 61.71 52.08 57.87 47.60 Small RFS's '1? cohesion 84.10 76.00 72.78 67.00 Large RFS's 3? cohesion 40.75 31.57 35.67 39.29 r8 between size and cohesion .88 .85 .65 .28 Although a large difference in the correlations appeared be- tween freshman and senior year, only a small mean cohesion score difference occurred. The freshman mean of 61.7 differs only 22 per- cent from the senior mean of 47.6, but in the Opposite direction from the hypothesized relationship of increased size and increased cohesion. Thus, apparently an inverse relationship exists between size and co- hesion, as stated by Kretch, Crutchfield, and Ballachy. 1Supra, p. 18 for a discussion of Gordon's findings. 99 It was also hypothesized, to correlate with Kretch, Crutchfield, and Ballachy, that the smaller RFS's would be more cohesive than the larger RFS's each year. The mean cohesion scores between large and small RFS's each year show some support for the hypothesis, although the percent difference was less in the twelfth grade than in the ninth, 41 percent and 52 percent respectively. This smaller percent- age reflects the skewed senior year scores previously substantiated by the low correlation between size and cohesion. The between year mean cohesion differences were approximately 50 percent for both years. Although a hypothesis was not deve10ped concerning the compara- tive difference of small group means between ninth and twelfth grades and large group means between ninth and twelfth grades, they were checked. A 20 percent difference appeared in the ninth and twelfth grade small group means, but only a 4 percent difference in the large group means. Therefore, most of the total difference between p0pu- 1ations was in the small groups, directly reflecting fewer small groups in the senior year and more out-group choices leading to lowered cohesion within the three and four person groups. Thus, the small groups' cohesion scores also tend to refute Hallworth's find- ings of increased group cohesion in a longitudinal perspective. 100 Change in Size Of Social Acceptance Categories Basic to this research is the difference in studies basing social acceptance on number Of choices and studies formulated around accept- ance based on reciprocated relationships creating structures. One factor automatically tending to refute high correlations in an Open social system is change in number Of respondents. School drOpouts in a relatively closed social system lead to some alteration in num- ber, unless of course a single classroom is studied. Most school systems establish a basic pupil—per-teacher system which creates relat- ively comparable numbers in a single classroom. Yet an assumption Of static numbers in any open social system seems unreasonable. The third set Of hypotheses was deve10ped to eXplore the stability Of size, or number Of respondents, in each reciprocal friend- ship category; isolates, mutual pairs, and RFS's. It was hypothesized that, given a small overall number, the number Of respondents in each reciprocal friendship category will exhibit little change over a four year time span. Table XIII contains a complete breakdown Of the cate- gories, giving raw numbers and percentages Of the 1962 category N's represented in each subsequent year. Part of the yearly differences are attributable to the varying population N's: 154, 155, 143, and 138 from ninth through twelfth grades respectively. Yet Table XIII reveals two changes in the between year data which cannot be attributed to shifting population N's. In the tenth grade the percentage in 101 mutual pairs and isolates shifted although the total population N varied only by one person from the previous year. The percentage Of isolates increased 17 percent in the tenth grade. The trend continued in the eleventh grade when 56 percent more isolates appeared although the pOpulation decreased from 154 to 143. This trend continued to support Gordon's finding of structural breakdown in his eleventh grade group.1 The theoretical idea Of changes in one sub-unit balancing changes in another is also empirically demonstrated by the decrease in RFS mem- bers which compensates for the increased number of isolates. Table XIII. Four year summary by number and percentage Of the size, or number Of occupants, in the reciprocal friendship categories among the girls in one high school senior class, 1965 Year and Grade 1962 1963 1964 1965 Ninth Tenth Eleventh Twelfth Structural Categories: Grade Grade Grade Grade (N) (N) (N) (N) Isolates 23 27 36 20 Mutual Pairs 18 16 18 16 RFS's 113 112 89 102 % Of 1962 N: Isolates 117 156 87 Mutual Pairs 89 100 89 RFS's 99 79 90 'Sugra, p. 18 for a discussion of Gordon's findings. 102 The data did not support Gordon's finding that twelfth grade isolates were fewer, having been absorbed in mutual pairs.1 When the decreased pOpulation N in the twelfth grade (from 154 to 138) is con- sidered, 20 isolates in the twelfth grade as compared tO 23 in the ninth grade does not constitute change. TO summarize, three changes in the number of occupants, or size of social acceptance categories, tend to refute the hypotheses which pre- dicted little change in number: isolates in the tenth and eleventh grades. and RFS members in the eleventh grade. Infiaddition, the decrease in population N's from the ninth to twelfth grades automatically introduces some change. Individual Choice Status vs. Reciprocated Relationships The above hypotheses were formulated considering interaction through reciprocation Of choices, whereas the refuted studies were based only on number Of choices. Therefore a fourth hypothesis was deve10ped to explore the difference in social acceptance as measured by choice status, and social acceptance based upon reciprocated choices, creat- ing structures. The hypothesis stated that the individual respondent's choice status excluding reciprocated choices, will not coincide with position in a structured group based upon reciprocated choices. Two checks were made tO see whether or not choices disregarding interaction coincided with choices including reciprocation. Figures 7 and 8 contahi the number of choices made tO each individual2 on the friendship ques- tion which by its design3 automatically establishes a reciprocity 1Supra, p. 18 for a discussion of Gordon's findings. 2Choice status is indicated by small numbers attached to each individual. 3Lines indicate reciprocated choices between individuals. ‘5’ ‘3’. 1: 112; 1'1 11 11 4 RESH4 o 030’ @394 :’e "10001 56 a 04 RFSAQS ‘ z 5.3 4. 519 m RFS/17" A “.9 0-- ® is RFS [/6 ”so G®@®’ 0’ E1 ’G G" .1 mm v - 5.9 IZS 087 K O \ m k k «-flc~anso -'ro~ru5to' -'S£ZF' v ISUKAIES 4‘ -0 a . HMMAEES - ' [$014723 '01?) v .2 4 @ RFSIDG RFS/08 Reciprocal Friendships of the Ninth Grade Girls in One High School Class Which Became the Senior Class of 1965 With Individual Choice Status Indicated for Each Girl Figure 7. uOflcom Hoonum nmflm 0:0 aw mauww moose numao3a on» no newsmoamwum Hmoouawoom O I {1 \\ OIC 1 . be» 133‘ \LI/K I18; IV1_.IK \. . , o .4, ....I Mmovmmf I. n08.” \ Vx‘ MKQ III/en III w /w I lu11} Inn: '1. 111‘ III, g In I ,I . 1.1 «FININKI ~,,In.vI.I vmwlv IpWWI Hose some now oouooaoau nausea ooeoeo Hoaoe>eooa noes meme mo nooao IL I e to: Mmuwfi 11.1.. “X I II \I.1 1 .EIIII .. to. 7. .6 V IF. .. IFN .I 11414151”. I/ VAN. L x I .. u I... 11-11,. seaside. - 31:36.3 use... .Ifiooi. :62 r A _/1111 J I. .. 3mm. 1 .1 we? 52 L .1 1 I, K. 1I can: I mo as o. .6. 1A..» n IA 4 J \1 l. “‘11 1r x ..I , I® InooI IaonI 0.. 1 0.4 1. o x . .1 I/A I11.I.) so I . .I 19249.6? 1 $3363 a: $0wa 30.. I...y I . I ._ L \R .1 k... v. ... .RILI \\ 11.11 FULL” *1COW in»! .r ImVOv on: NVOVI. Am: ..I 0.1 A 0v. 3. Ci. ..IA 0 I .I! NY. IJA. 4 11 _ 14 d .133. 1 42:. so? _ I .YI v [x and: NON ..TIIIIWOnoIIv n 1 IA. If. .. J. 4 hIIl.) \I .\.1/ ..I .I «a . I .. .I we.» eke one. 1.. .3 sm no\lI.n~oI 3:3 #333 -..\ I . I 1W\ « .I., ammonlm ooo.r1l~..,.m3 I 1 v/II 1.1‘ a H NOV WKQ .x. r I ‘II IIoImI IIemoviInfiMoo (1 \ \‘ 11K .\ ,. K .r I“ . V I I. I .33 :mIoT .Imnolmglmmo 11 . N I K .. fill 1. awe/,1 . I I I, /1 I In. ..I . ~0V_.IKKVW1\Q AV: . A40 ..II C/ k e . ..I open: Tl IecoI n... A L ..._ I L. omo. o: .1 £8; ~ 1 1‘ n 1,1.1A 0 I 1, Nn‘ hkk I . 1r. .1 A, 1/ tin. /A.I . 5911002113? v z, ./.I v 71.1. NTK 3...: o no .335 :6: .Iao 30:52 .0 Loan: col-Sud So :34 So?! 3304 1- x I axc n. I If I. mmduu 4Hw mos mmm comm .ousuoouum dwempaowuw amoounwoou Ou muomou mmm ”name Hmouse ou muomou m2 moumHOmfi ou muomou H .onhu mono: mam .moum wcwaaosm .aowumaoo noo no momma wcwumu mmmHo HmmOOm m .mowumwuouooumeo maumum mo woocH mHNOOHmz Ou wuomou UmH "ammo AooHv amH m m m m N m a e m a NN a a m e mH m mH mN nHmsom HoHV 6H OH H N H H N m em-oN anaHu N H H me-Ne HaHeom a H N H ee-Nm asaoH Houdm mmuam omumm mmoao Hmwoom mapofiz mH H H m H a H H H N a Nm-am on N H H e H H m N H N H a m e a N mm-mN annHo H H aN-mN HnHeom NH H H H H H m N H NN-mH mama: N H m H H sH NH-NH a z NHH eHH mHH aHH mHH NHH HHH oHH moH moH NoH eoH moH aoH moH NoH HoH nwsHsaa nHmsoa mam mmm mmm mmm mmm mmm mmm mmm mmm mmm was mmm mmm mmm mam mmm was ex. H pmsemHsz nanHo mood .mmmHo Howcom on» madden nowsz mmmao Hoocom swan oao ca mauwm okuw mucus Noma onu waofim mwwuomOONO mocmunooom amwOOm an cOHusnwuumfip mmmHO gmHOOm Hmowumadz momH Hmwoom .H sHsme All Evin 3. fl 1. din. l , inv n 196 .huowoumo some cw mfiuww mo Honaszn .cowu nmowwwucovfi mom women: m co>Hw mos mum zoom .OHOHOOHum dwemooowum Hmooudwoou OH muowou mmm “Name Hmsusa OH muomou m2 moumHOOH Ou muomou H .6dhu omso: mom .moum mafiaaoav .cowumnno 100 do momma wnwumu mmmao HmNOOO m .mowumwuouomuoso msumum mo xopOH m.uocum3 ou mummou omH "homo HOOHO NNH N N N 4 NH N NH N N NH O ON s OH NN anaoa HNO NH O a H N H ON-ON nanHO H H OO-NO HnHoom a H H H H OO-NO assOH N H H HO-aN manO N H N NN-HN HaHeom Na H H N N N N NH N NH ON-NN aHOOHz GB.....ms.6.,......h.,. ,w . , ,NH .H N H N N a a NN-aN NN H N N H H a N N O 6 NH N N NN-NN nanHO N N H H N H H «N-NN HsHsom OH H H a N H NN-NH neaaO O N H sN NH-NH N z ONN NHN OHN aON NON NNN NON NNN NHN NHN NNN NON HON emsHHsm nHmaoa Nam mam Nam Nam Nam NHN NHN mmm mmm NNN mam. mam mam H: N OsaemHas anaHO moma ammoao uoHcom oeu mamoon moans mmmfio Hoosom swan Odo cw mduww oomuw eucOH mama onu maofim wowuowouoo oocmuaooom Hmaooo he cOHuanNuumHv mmmao HmfiOOn HOONHOESZ .N oHan .muowoumo some ON mauww mo wonadzn .OONu umowmwucopH-u0m Homes: 6 =o>Hw was mm“ somm .ou5uonuum naemvcowum Hmoounwoou OH muomou mum Huwmd Humane OH muomou mz mouoHOOH OH muomou H .OONO mason poo .soum woNHHoam.aoNuonoo 100 no momma wcwumn mmmao Hmwoow m .mowumwuouomumno maumum mo xoch m.uoaum3 Om muowou UmH "home HOOHO NOH O N N a N N N N NH O N HH O NH N 197 NH ON NHOOON HNO N a H H H H sm-ON mmmHO H H OO-NO HmHuom N H H OO-NO sssoH a H H H H H0-0N smaHO O H H N NN-HN HnHoom NN H H H N N N N N N H O O ON-ON sHOOHz NNOO,.NO.-:-:.-:.V . NH H H H N H H H H N NN-ON NN N N N N N N N N N N O H N HH NN-NN manO N H N ON-NN HmHeom OH H N H H H N N NN-NH seam: O H N ON NH-NH N z NHN NHN ONN NNN NNN ONN NNN NON NON NON OON NNN NON NNN NON mmOHOam wfimuoa mhm mmm mhm mmm mmm mmm mmm mmm mmm mmm mm“ mmm mmm mhm mmm m: H pounwwmz mmmau mumH dmwoom moma .mmmHo NONGOO ego oaooon soana woman #00506 cm“: Ono cw mauww ovmhw zuco>oHo doma Ono woofim mowuowoumo oocmuaooom HONOOO he cofiuanwuumwv mmmao HOHOOO Hmowuoaaz .m OHDOH 198 .zuowOHOO some OH mHNHw Ho Nonadzn .ooHu nOOHHHuaopH How Hones: m no>Hw mmz mmm zoom .ouOuosuum OHemvcoHuH HmoounHoou OH muowou wmm muwmn Hanuoa OH muomou Hz ”OHOHOOH ou muomou H .mnhu omso: mam .moum wzHHHosp .coHumnso 100 co momma wcHumu wmmHo HOHOOO m .mowumHu6uomumeo mdumum Ho xOvOH m.uocum3 OH mHoHouNUmH "moms HOOHO NNH ON N O N O NH N N ON N OH ON OHOOON HOO, N N H H H Om-oN mmmHo H H OO-NO HmHOON H H OO-NO umsOH Lllll % NNNO NO , N H H H0-0N mmnHO N H H H NN-HN HmHoom OO N H H H O N N H O H N N ON-ON OHOOHz ibw is. 1. NH O N H H N N NN-ON NO NH N H N O N N OH O O NN-NN mmaHO N H H ON-NN HaHoom mH H H H H O m em mmuwH Home: NH N O H N NH-NH N z NH-NHO OHO OOO NNO NNO NOO NNO NNO NNO HOO swaHsmm mHmHON NON Nae NNN NHN Nam Nam NNN NHN NON was as H Omsemes nasHO mUmH HmHoow ll moaH .mmmHo uoHcom Hoonom an5 OOO OH mHun ovmuw :uHHosu moaH onu wooem OOHuoonmo ooamudooom HOHOOO >9 coHuonHuumHv mmmHo HmHOOO HOOHuoEdz .d oHan 199 .OONO omno: was .moum wcHHHosv .coHummso uoo co momma wcHumu mmmHo HmHOOO m .mOHuOHuouomuoso maumum mo xomcH m.uo:um3 ou muomou,omH m HONHOOOH O N NH NN OO N O HONHOOOH H OH OH ON OH HH N HONHOOOH H O O O NH ON NO OHNHO HHO HNHHOOOH N N HH NN HO N HHHNHHOOOH - N OH ON NH OH O HNHHOOOH - N O O NH ON NO assess: OOO HOHO OOH O O OH NO NN O - HOHO OOH O NN NN ON NH - - HOHO OOH O - NN O NH NN NN Nassau: HHOH Hanan: NNNO OOH O N OH NH HO N O NNNO OOH N O NH HO NN N N NNNO OOH - OH - N OH NN ON nassHOOH sz N N N N N N N N sz N N N N N N N N sz N N N N N N . N OOHuowoumo OHOOOH N c n O m N H OHOHOH N c n O m N OHOOOH N o n O m H dHnmpOOHum Hmoouawoom ...: wdHumm swam MOHHHosa ,wdHumm OONH Omsom ,MMHOOM HmcoHummmooo nHNHO OOOHO euaHz HON OOOHOOH OOH HOOOH>HOOH momH .ommHo HOHOOO oeu mamoon LOH53 OmmHo Hoosom ewH: Odo OH mHun ovmuw nucH: NomH ocu macaw OOHuow -ouwo nHevaOHum HOOONOHOON new OOOHOOH momH HmspH>HvsH OH waHpuooom OOHOOHHHHONO owmucmouom .m OHHOH 200 .OONH omson .00 co momma wcHumu mOmHo HmHoom m .OOHHOHHmuomHmnu maumum Ho mam .moum wcHHHoav .coHumdoo xomcH m.uoaum3 OH OHOHOH UmHm NNNHOOOH O N NH ON ON O HHNNNHOOOH H O OH HO ON NH NNNNHOOOH H O O N OH ON HO OHNHO HHO NNHHOOOH N N HH ON ON N NHHNHHOOOH H O NH ON NN NH NHNHHOOOH H O O O NH NN OO nesasmz OOO HOHO OOH - O OH OH OO O O NOHOOOH - - NH ON - NH -NOHO OOH - O - OH NN HN OH masses: HHmm Hanan: NNNO OOH - O OH ON NN O N NNNOOOH O - ON OO OH O -NNNO OOH O - O - OO ON ON nsaaHoOH sz N N N N N N N N H2O N N N N N N N N H2O N N N N N N N N OOHHOOOOOO OHOOOH N O N O N N H OHOOOH N O N O N N H NHOHON N O N O N N H aHemOsaHam HmooumHoom wdHumm mou<,wdHHHosa wdHumm omNH undo: wdHumm HmcoHummwooo NHHHO OOOHO enema OOH neeHOOH OOH HOOOH>HOOH momH .mmmHo MONGOO mam oamoon eOHes mmmHo Hoonom :mH: OOO OH mHun onmum :uzou momH Ono waoam OOHuowoumo OHnmpcmHHH HOOOHOHOOH mam OOOHOOH mumH HmsmH>HmGH OH wcwpuooom coHuonHHumHm owmunoouom .o oHan 201 .OONO omso: mam .moum wcHHHoam .OOHHOQOO 100 co momma waHumu OOmHo HOHOOO m .OOHOOHHouomHmeo maumum Ho xomcH o.uo:um3 ou OHOHOH.UmHm HNOHOOOH - N ON NO ON NH NHNOHOOOH O N O NN OO OH NHNNOHOOOH H N N N OH ON ON nHNHO HHO NOOO OOH - O NH NN NN OH HNOOO OOH H N O ON HO OH OHHOOO OOH H H N O OH ON OO Nassau: OHO HOHO OOH - HH ON OO NH - -HOHO OOHNH N NH ON ON N N HOHO OOH - N HH O ON NN NH Nassau: HHOH Hanan: NONO OOH - - NN HN ON OH NHONO OOH N - HH NH NO HH HHHONO OOH - N N N NN ON ON nsOnHeaH H2O N N N N N N N N HzO N N N N N N N N HzO N N N N N N N N OOHNOOOOOO OHOOOH N O N O N N H OHOOOH N O N O N N H OHOOON N O N O N N H OHOOOOOHHO 1 . Hmooudwoom ,onumm mou< wcHHHozn, wcHumm OQNH,OO=om ,wdHumM HmcoHumasooo nHNHO OOONO OHOO>OHO HOO OOOHOOH OOH HOOOH>HOOH mooH .OOOHO HOHOOO osu osmoon LOHLB wOmHo Hooeom :wHe Ono OH OHHHw ommuw nuco>oHo OomH oeu wcoam mowuowoumo stOvOOHHH HmoounHoon pom OOOHOOH HOumH HmOvH>chH Ou wchuooom coHuonHuumHm.owoOcoouom .N OHan 202 umdsooo no momma wcHOmH OOOHO HOHOOO m .OONO omoo: mom .moum wcHHHosm .coHu .OOHOOHHOOOOHOLU maumum Ho xoch O.uocum3 Ou OHOHOH umHm NONHOOOH N N O HN OO O NH HONHOOOH - O ON NN ON NH O HONHOOOH H N N N OH ON NO OHNHO HHO HNOHOOOH N .N.,O NN NO O ON HNHHOOOH - N NH NN NN NH O NNOHOOOH H N N N NH NN OO assess: ONN NOHO OOH - - O OH ON NH O HOHO OOH - - OH HN HN OH - NOHO OOH - - - O NH ON HN Oneessz HHOO Hanna: HONO OOH N - ON NH NO NH - HONO OOH - N NN NN ON N - HONO OOH - N - OH ON NN OO ssaaHOOH NzO N N N N N N N N HzO N N N N N N N N HzO N N N N N N N N naHneOsOsO OHOOOH N O N O N N H OHOOOH N O N O N N H OHOOOH N O N O N N H OHOOOOOHNO HmooumHoom OOHOOO nasO OOHHHmso ,wcHumm OONH,OOOOm ,waHumm HOOOHummsooo OHNHO OOONO OOOHase HOO OOOHOOH OOH HOOOH>HOOH momH .OOOHO HOHOOO Hooeom ewHe OOO OH OHHHw Ovmuw nuHHosu momH oeu wcoam OOHuowoOOo OHLOOOOHHH HOOONOHOOH mam OOOHOGH momH HmovH>chH OH wchHooom soHuonHuumHv owmucoonom .w OHHOH . 203 Table 9. Individual sociometric pattern for the girls in one high school senior class, 1965 Year and Grade 1962 1963 1964 1965 Sociometric Ninth Tenth Eleventh Twelfth Pattern Number Grade Grade Grade Grade (N=154) (N=155) (N=143) (N=138) (Total N=l96) 001* RFSa 104 RFS 204 RFS MP RFS 432 RFS* 002* RFS 104 RFS 204 I1 RFS 417-13 A* 003 RFS MP - - - 004 RFS MP I2 - - a 005 RFS 104 - - - 006 RFS 104 - - - 007 I2 RFS MP RFS 350 - A} 008* RFS 108 RFS 213 11 11 .43* 009* RFS 113 RFS 220 RFS 333 RFS 417-13 RFS* 010 RFS 108 RFS 213 I4 - A: 011 RFS 108 RFS 220 - - RFS 012* RFS 102 RFS 202 RFS MP RFS 432 RFS* 013 RFS 108 - - - 014 RFS MP - - - 015 RFS MP 13 - - A 016 RFS 107 RFS 227 RFS MP RFS 427 RFS* 017* I3 RFS 213 I4 RFS 432 4* aKey: RFS = reciprocal friendship structure Dal-1H mutual pair isolate (subscript not in pOpulation during year in pOpulation for all four years change in pattern = class of I) Table 9--Continued 204 Year and Grade 1962 1963 1964 1965 Sociometric Ninth Tenth Eleventh Twelfth Pattern Number Grade Grade Grade Grade (N=154) (N=155) (N=143) (N=138) (Total N=l96) 018* 13 RFS 213 RFS 332 RFS 432 23* 019* RFS 107 RFS 207 RFS 307 RFS 407 RFS* (7) 020 RFS 107 RFS 219 RFS 332 - RFS 021 RFS MP - - - 022* RFS 112 RFS 217 RFS 317 RFS 417-13 RFS* 023* RFS 112 RFS 205 I2 RFS 460 L)* 024 RFS 112 RFS MP - - RFS 025* RFS 112 RFS 205 RFS 335 RFS MP RFS* 026* RFS 111 13 I3 RFS 427 AA*- 027* RFS 111 RFS MP RFS 332 RFS 452 RFS* 028* RFS 110 RFS 205 RFS 335 RFS 432 RFS* 029 RFS 110 I3 - 12 43 030* RFS 110 I4 I2 RFS MP 13* 031* RFS MP RFS 205 RFS 335 RFS MP RFS* 032 RFS MP 14 RFS MP - 4> 033* RFS 104 RFS 204 RFS MP RFS 432 RFS* 034 RFS lll RFS 207 - - RFS 035* RFS 103 RFS 202 RFS 307 RFS 407 RFS* 036* RFS 107 RFS 237 RFS 337 RFS 407 RFS* (7) 037* RFS 111 RFS 207 I4 RFS 407 A* 038* RFS 107 RFS 207 RFS 367 RFS 407 RFS* (7) Table 9--Continued 205 Year and Grade 1962 1963 1964 1965 Sociometric Ninth Tenth Eleventh Twelfth Pattern Number Grade Grade Grade Grade (N=154) (N=155) (N=143) (N=138) (Total N=l96) 039* RFS 107 RFS MP RFS 307 RFS 407 RFS* 040* 12 RFS 202 RFS 332 RFS 407 43 * 041 RFS 107 RFS MP - - RFS 042* RFS 107 RFS MP RFS 347 RFS MP RPs* 043* RFS 107 RFS 207 RFS 307 11 <3 * 044* RFS 102 RFS 202 RFS MP RFS 432 RFS* 045 RFS 102 - - _ 046* RFS 102 RFS 202 RFS MP I4 2d'* 047* RFS 102 RFS 202 RFS MP RFS 401 RFS* 048* RFS 102 RFS MP RFS MP RFS 401 RFS* 049 RFS MP - - RFS 417-13 RFS 050* RFS MP RFS 217 RFS 317 RFS 417-l3 RFS* 051 RFS 109 - - - 052 RFS 109 RFS 213 - - RFS 053 RFS 109 - - _ 054 RFS 109 — - - 055* RFS 109 RFS 202 RFS 322 RFS 417-l3 RFS* 056* RFS 117 RFS 217 RFS 317 RFS 417-13 RFs*(l7) 057 RFS 109 RFS 202 RFS 322 - RFS 058* 14 RFS 217 RFS 330 RFS 417-13 2>.* 059* RFS ll7 RFS 217 RFS 317 RFS 417-13 RFS*(17) Table 9--Continued 206 Year and Grade 1962 1963 1964 1965 Sociometric Ninth Tenth Eleventh Twelfth Pattern Number Grade Grade Grade Grade (N=154) (N=155) (N=143) (N=138) (Total N=l96) 060* RFS 102 RPS 222 RPS 322 RPS 422 RPS* (2) 061* RPS 102 RPS 222 RPS 322 RPS 422 RPS* (2) 062* RPS 102 RFS 202 RPS 322 RPS 401 RPS* 063* RFS 102 RPS 202 I1 I2 4 * 064 RPS 103 I2 - - 23 065* RPS 113 RFS 213 RPS 333 RPS 417-13 RPS* 066* RPS 113 RFS MP 12 RFS 452 «A * 067* RPS 113 RFS 213 RPS 313 RPS 417-13 RPS* 068* RPS 115 RPS 202 RPS 322 RPS 432 RPS* 069* RPS 113 RPS 213 RPS 313 RPS 417-l3 RPS* 070* RPS 113 RPS 213 RPS 313 RPS 417-13 RPS* 071* RPS 110 RPS 205 RPS 330 RPS 417-13 RPS* 072 RPS 110 I2 - - ‘4 073* RPS 110 I4 RPS 305 RPS MP 4‘* 074* RPS 102 RPS 213 RPS 342 I3 43* 075* RPS 102 RPS 222 I3 RPS MP ~4* 076* RPS 102 RPS 213 RPS 342 I4 4* 077* RPS 102 RPS 213 RPS 342 14 A* 078* RPS 113 RPS 213 RPS 313 RPS 417-13 RPS* 079* I2 I2 12 12 I* 080* RPS 112 RPS 204 I1 RPS 432 43* 081* RPS 103 RPS 202 RPS 332 RPS 432 RPS* 207 Table 9--Continued Year and Grade 1962 1963 1964 1965 Sociometric Ninth Tenth Eleventh Twelfth Pattern Number Grade Grade Grade Grade (N=154) (N=155) (N=143) (N=138) (Tatal N=l96) 082* RFS 103 RFS 202 RFS 332 RFS 432 RFS* 083* RFS 103 RFS 202 RFS 332 RFS 432 RFS* 084* RFS 106 RFS 202 RFS 332 RFS 432 RFS* 085 RFS 103 RFS 202 RFS 332 - RFS 086* RFS 107 RFS 202 RFS 307 RFS 477 RFS* 087* RFS 107 RFS 202 RFS 307 RFS 477 RFS* 088* RFS 107 RFS 202 RFS 307 RFS 477 RFS* 089* RFS 107 RFS 207 RFS 367 RFS 407 RFS* (7) 090* RFS 106 RFS 202 RFS 322 RFS 432 RFS* 091* RFS 107 RFS 222 RFS 347 RFS MP RFS* 092 RFS 107 - - - 093* RFS 107 RFS 222 RFS 347 RFS 407 RFS* 094* RFS 107 RFS 227 RFS MP RFS 427 RFS* 095* RFS 107 12 12 RFS 427 A * 096* 12 RPS 220 I4 RPS 417-l3 CH 097 I3 - - - 098 RFS 106 RFS 213 - - 099* RFS MP RFS 213 RFS 333 RFS 417-13 RFS* 100* I4 RPS 202 RPS 332 RPS 432 A * 101 RFS 105 RFS 205 - - RFS 102 RFS 105 RFS 205 12 - 43 103 RFS 105 RFS 205 RFS 305 - RFS (5) 208 Table 9--Continued Year and Grade 1962 1963 1964 1965 Sociometric Ninth Tenth Eleventh Twelfth Pattern Number Grade Grade Grade Grade (N=154) (N=155) (N=143) (N=138) (Total N=l96) 104* RFS 105 RFS 205 RFS 305 I4 4* 105* I2 RPS 205 I4 I2 43* 106* RFS 101 RPS 201 RPS MP RFS 401 RFS* 107 RFS 101 - - - 108* RPS 101 RPS 201 RPS MP RFS 401 RPS* 109 14 RFS 201 - - 23 110 I4 RPS 201 - - A: 111 RFS MP RFS 207 - - RFS 112 12 12 - — 1 113* I3 RPS MP RPS 332 RPS 452 *9 * 114* RPS MP RFS 237 RPS 337 RPS 407 RFS* 115* RPS MP RPS 207 RPS 307 RPS 477 RFS* 116 RPS MP RPS 237 RPS 337 - RPS 117* RFS MP RPS MP 13 12 A * 118* RPS 107 RPS 202 RPS 307 RFS 477 RFS* 119* RFS 107 RPS MP RFS 347 RPS MP RPS* 120* I2 RFS 202 RPS 307 RPS 407 A * 121* RPS 107 RPS 237 RPS 337 RPS 407 RPS* (7) 122* RPS 107 RPS MP RPS 307 RPS 407 RPS* 123* RPS 107 RPS MP RFS 307 14 ll-* 124 RFS 107 RPS 207 - - RPS 125* RPS 107 RPS 207 RPS 307 RFS 477 RFS* (7) 209 Table 9--Continued Year and Grade 1962 1963 1964 1965 Sociometric Ninth Tenth Eleventh Twelfth Pattern Number Grade Grade Grade Grade (N=154) (N=155) (N=143) (N=138) (Total N=l96) 126* RFS 107 RPS 207 RFS 307 RPS 407 RFS* (7) 127 I2 RPS 202 RPS 340 - £5 128 I1 - - - 129* RPS 115 RPS 202 RPS 322 RPS 432 RFS* 130* RFS 106 RPS 202 RPS 332 RPS 414 RPS* 131* RPS 104 I4 RPS MP RFS 432 .23* 132 12 - - - 133* RFS 104 14 I1 I3 {5‘* 134* RPS ll4 RPS 202 RPS 332 RPS 414 RFS* 135* RPS 114 RPS 202 RPS 332 RFS 414 RFS* 136* RPS 114 RFS 213 RFS 342 RPS 432 RFS* 137 11 RPS 213 - - ‘3 138 RFS 102 RPS 202 - - RPS 139* RPS 109 RPS 202 RFS 322 RPS 422 RFS* 140 RPS 109 - - - 141* RPS 115 RPS 202 RPS 322 RPS 432 RPS* 142* RPS 107 RPS 219 I4 RPS 422 ‘3'* 143* RPS 116 RFS 216 I4 RFS 432 A * 144* RPS 116 RPS 216 RPS MP RFS MP RFS* '145* RFS MP RFS 205 I3 I3 43'* 146* RPS MP I3 I1 11 Av. 147* I2 RPS 202 I4 RPS 432 A * 210 Table 92-Continued Year and Grade 1962 1963 1964 1965 Sociometric Ninth Tenth Eleventh Twelfth Pattern Number Grade Grade Grade Grade (N=154) (N=155) (N=143) (N=138) (Total N=l96) 148* RFS 117 I3 RPS 317 RFS 417-13 A* 149* 13 13 RPS 317 RPS 417-13 41=* 150 RPS MP - - - 151* 14 RPS 202 RPS MP RPS 432 A'* 152 RPS 116 - 12 I4 Ab 153* RPS 104 12 RPS 33o RPS 417-13 43=* 154 12 - - - 201 RPS MP 12 RPS MP ‘3 202 I2 - - 203 RPS MP - - 204 RPS MP RPS 350 RPS 417-13 RPS 205 RPS 237 RPS 337 RPS 407 RPS (7) 206 RPS 207 RPS 307 RPS 477 RPS (7) 207 RPS 205 I3 RPS MP £1 208 12 I3 RPS 407 43 209 12 I3 RPS MP é§ 210 13 RPS MP RPS MP £3 211 13 RPS 350 13 13 212 RPS 216 - - 213 12 - - 214 RPS 219 RPS 232 RPS 432 RPS Table 9--Continued 211 m Year and Grade 1962 1963 1964 1965 Sociometric Ninth Tenth Eleventh Twelfth Pattern Number Grade Grade Grade Grade (N=154) (N=155) (N=143) (N=138) (Total N=l96) 215 RFS 213 I3 RFS 432 216 RFS 222 RFS 322 RFS 422 RFS (2) 217 14 - — 218 RFS 202 RFS 340 RFS 432 RFS 219 11 " - 220 RFS 227 I4 12 13 221 12 - - 301 RFS 332 RFS 452 RFS 302 RFS 340 - 303 11 RFS MP 43 304 12 - 305 RFS 335 RFS 432 RFS 306 RFS 313 RFS 432 RFS 307 RFS 332 RFS 452 RFS 308 RFS MP RFS 417~l3 RFS 309 RFS MP RFS 4l7~13 RFS 310 RFS 367 - 311 I3 - 312 I1 - 313 12 12 I 212 Table 9-—Continued Year and Grade 1962 1963 1964 1965 Sociometric Ninth Tenth Eleventh TWelfth Pattern Number Grade Grade Grade Grade (N=154) (N=155) (N=143) (N=138) (Total N=l96) 402 RFS 460 403 RFS 460 404 RFS 417-13 405 RFS 460 406 RFS 407 407 RFS 432 409 RFS MP 410 RFS MP Table 10. 213 Numerical comparison of reciprocated choices and choice Statuses among the girls in one 1962 ninth grade high school class which became the senior class, 1965 ReSpondent Numb er Friendship Question POpularity and Dress Questions RCa CS GS, MP CS,BD CS,NDR 001 002 003 004 005 006 007 008 009 010 011 012 013 014 015 016 017 018 019 020 021 022 023 024 025 026 027 028 029 030 031 032 033 034 035 036 NHHHHwaNHNHNHNv—INt—IOONHr—IHHHNr—INOHNHHNM l 9 l4 U1HwNwawaonHNHwt-awNNHNwmmwaOt-‘wr—‘Hwb H aKey: RC,FQ CS,PQ CS,MP CS,BD CS,NDR Number of reciprocated choices, or ties, friendship question. Choice status, or number of choices received, friendship question. Choice status, "most p0pu1ar" question. Choice status, ”best dressed." Choice status, "not dressed right.” 214 Table 10--Continued W Respondent Friendship Question Popularity and Dress Questions Number RC CS CS,MP CS,BD CS,NDR 037 038 039 040 041 042 043 044 045 046 047 048 049 050 051 052 053 054 055 056 057 058 059 060 061 062 063 064 065 066 067 068 069 070 07l 072 073 074 075 076 077 078 079 080 081 082 083 1 l4 4 10 1 23 7 16 4 8 9 23 25 b-l-‘wNOHHwDNwHDHbr—tr—iNNHNNwwNOHNww-PNwHHHwaHw-PNOCJNN H bbWDONNwm-PMr-‘O‘bciNMNNWNDLHbNHNNmeO‘LfiHHHUJN-DbmflNOW-Pb) 215 Table 10--Continued Respondent Friendship Question POpularity and Dress Questions Number RC CS CS,MP CS,BD CS,NDR 084 085 086 087 088 089 090 091 092 093 094 095 096 097 098 099 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 16 ll 16 16 N9 N w \IwOWHN H-l-‘UINUIOI-I r—INoowmbme'IOr-JHHHHHOOHCONNNONNHwOHHOOP-INHNNuwNwNwH H U1NOOJ-‘NankOO‘OUJHD-‘HNWWOHNHMPNOP-L‘H-PHNNHOHWHNPWPVWWDH 216 Table 10--Continued Respondent Friendship Question POpularity and Dress Questions Number RC CS CS,MP CS,BD CS,NDR 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 H OHNOHONOHHNNNHmwNONan—‘Ob H OHNbNNNONNNNwwmme-Pw-L‘NOW Table 11. 217 Numerical comparison of reciprocated choices and choice statuses among the girls in one 1965 twelfth grade high school senior class, 1965 Respondent Friendship Question POpularity and Dress Questions Number RCa CS CS,MP CS,BD CS,NDR 001 2 4 002 2 3 003 004 005 006 007 008 0 0 2 009 1 2 010 011 012 3 4 013 014 015 016 3 3 1 017 3 3 018 4 9 1 1 019 l 1 5 020 021 022 4 4 023 l 3 024 025 l l 026 1 3 027 3 4 l 028 2 2 029 0 0 030 1 1 031 1 2 032 033 2 4 034 035 2 5 l l 036 3 6 4 1 Key: RC,FQ = Number of reciprocated choices, or ties, friend- ship question. CS,FQ = Choice status, or number of choices received, friendship question. CS,MP = Choice status, "most popular" question. CS,BD = Choice status, "best dressed." CS,NDR = Choice status, "not dressed right." 218 Table llv-Continued Respondent Friendship Question Popularity and Dress Questions Number RC CS CS,MP CS,BD CS,NDR 037 1 l 1 4 038 2 5 l 3 039 2 3 13 21 l 040 1 1 041 042 1 4 3 12 043 0 0 8 15 1 044 2 4 045 046 0 l l 047 2 3 048 3 3 049 2 3 050 3 3 051 052 053 054 055 l 2 056 2 3 057 058 2 3 059 4 4 060 4 4 l 061 2 3 062 1 2 063 0 0 1 064 065 3 4 8 066 1 2 067 3 3 068 2 5 069 4 4 l 070 3 3 071 4 5 072 073 1 4 074 0 1 075 l 5 l 076 0 l 077 0 1 078 4 6 079 0 0 080 2 3 081 3 6 l 082 l 2 2 083 3 8 l 4 084 3 5 l 085 219 Table ll--Continued Respondent Friendship Question Popularity and Dress Questions Number RC CS CS,MP CS,BD CS,NDR 086 087 088 089 090 091 092 093 094 095 096 097 098 099 l 100 3 4 101 102 103 104 105 106 107 108 2 2 7 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 3 130 131 3 132 133 0 4 l P‘P‘UJP‘UJNJ NH-L‘LDUDN 39 3 23 3 wrap-1H 011—nub |'—' N NCO NO!-‘ NO‘ HWN LAD-DUO boo NN OHNNr—Iwo tow NJ-‘J-‘LANUJO H H H Drip.) gnu-Ru ,. - 220 Table 11--Continued Respondent Friendship Question Popularity and Dress Questions Number RC CS CS,MP CS,BD CS,NDR 134 1 2 135 2 4 2 2 136 l 3 4 1 137 138 139 2 6 140 141 2 2 142 1 2 143 2 3 144 1 4 145 0 2 146 0 0 3 147 2 4 148 1 1 149 6 6 4 150 151 3 10 152 0 1 153 1 1 l 154 201 1 1 1 204 2 4 205 4 5 206 2 2 207 1 3 2 208 4 5 2 l 209 1 1 210 0 4 214 1 1 2 215 3 3 216 1 1 1 1 218 5 5 220 0 0 l l 301 3 3 1 303 l l 305 2 2 306 1 1 307 3 4 308 4 5 l 309 4 5 313 0 0 1 Table ll--Continued 221 W Respondent Number Friendship Question POpularity and Dress Questions RC cs CS,MP CS,BD CS,NDR 402 403 404 405 406 407 409 410 I—‘I—‘NNNNHN HH-DwNwP—tw B IBLI OGRAPHY Ea BIBLIOGRAPHY Barber, Bernard. Social Stratification. New York: Harcourt, Brace, and WOrld, Inc., 1957. Becker, Howard and Useem, Ruth H. "Sociological Analysis of the Dyad," American Sociological Review, VII (1942), 13-26. Bjernstedt, Ake. Interpretations of Sociometric Choice Status. Lund, Sweden: CWK Gleerup, 1956. Bjorngaard, Arlene. "The Relationship of Social Class and Social Ac- ceptance to Clothing and Appearance of a Selected Group of Ninth Grade Girls," Unpublished Master's thesis, Michigan State Un- iversity, 1962. Blair, William Myers. Elementary Statistics. New York: Henry Holt Co., 1952. Brookover, Wilbur B., Smucker, Orden C., and Thaden, John Fred. Soc- iology of Education. New York: American Book Company, 1955. Cannon, Kenneth L., Staples, Ruth, and Carlson, Irene. "Personal Appearance as a Factor in Social Acceptance," Journal of Home Economics, XLIV (1952), 710-713. Cartwright, Darwin and Zander, Alvin. "Introduction: Group Cohesive- ness," Group Dynamics, Research and Theory. Cartwright, Darwin and Zander, Alvin, eds. Evanston, 111.: Row Petersen and Co., 1960, 69-75. Coleman, James S. The Adolescent Society. New York: Free Press, 1961. Cook, Lloyd Allen. "An Experimental Sociographic Study of a Stratified 10th Grade Class," American Sociological Review, X (1945), 250-261. Davis, Allison. "Socialization and Adolescent Personality," Readings in Social Psychology. New York: Henry Holt and Co., 1952, 520-531. Davis, Kingsley. "The Sociology of Parent Youth Conflict," American Sociological Review, V (1940), 523-532. Davis, Kingsley. Human Society. New York: The Macmillan Co., 1949. Durkheim, Emile. The Rules of Sociological Method, G.E.G. CBtlin, ed., Chicago, 111.: University of Chicago Press, 1938. 223 ‘11 w. 224 Eicher, Joanne B. 1"Adolescent Girls' VieWpoints from Ninth Through Twelfth Grade Concerning Dress, Social Acceptance,and Related Factors," Unpublished research in process for Michigan State University Agricultural Experiment Station. Elkin, Frederick and Westley, William A. "The Myth of Adolescent Culture," American Sociological Review, XX (1955), 680-684. Farley, Reynolds. "Suburban Persistence," Principles of Sociology, Young, Kimball and Mack, Raymond, eds., New York: American Book Co., 1965, 228-239. Good, Carter V. and Scates , Douglas E. Methods of Research. New York: Appleton-Century-Crofts, Inc., 1954. Goode, William and Hatt, Paul. Methods in Social Research. New YOrk: McGraw-Hill Book Company, 1952. Gordon, C. Wayne. The Social System of the High School. Glencoe, 111.: The Free Press, 1957. Gottlieb, David and Reeves, Jon. Adolescent Behavior in Urban Areas. East Lansing, Michigan: Bureau of Research and Publications, Michigan State University, College of Education and Extension, 1962. Hallworth, H. J. "Sociometric Relationships Among Grammar School Boys and Girls Between the Ages of Eleven and Sixteen Years," Sociometry, XVI (1953), 39-70. Harary,Frank and Norman,Robert X. Graph Theory as a Mathematical Nbdel in Social Science. Ann Arbor, Mich.: University of Michigan, 1953. Hays, William L. Statistics for Psychologists. New York: Holt, Rinehart, and Winston, 1963. Hendricks, Suzanne. "Opinions on Clothing and Appearance as Related to Group and Non-group Membership of Twelfth Grade Girls," Unpublished Master's thesis, Michigan State University, 1965. Hollingshead, August B. Elmtown's Youth. New York: John Wiley and Sons, Inc., 1949. Homans, George. The Human Group. New York: Harcourt, Brace and Com- pany, 1950. Jennings, Helen. Leadership in Isolation. New York: Longmans, Green Co., 1950. Joselyn, Irene M. "The Older Adolescent," Values and Ideals of American Youth. Eli Ginsberg, editor. New York: Columbia University Press, 1961. 225 Kahl, Joseph A. The American Class Structure. New York: Rinehart and Company, Inc., 1957. Kretch, D., Crutchfield, R., and Ballachy, E. Individual in Society. New York: McGraw-Hill Book Company, Inc., 1962. Lazersfeld, P. and Barton. "Qualitative Measurement in the Social Sciences: Classification, Typologies and Indices." The Policy Sciences, Recent Deve10pments in Sc0pe and Method. Lerner, D. and Lasswell, eds., Stanford, Calif.: Stanford University Press, 1951. Leavett, Harold J. "Some Effects of Certain Communication Patterns on Group Performance," Readings in Social Psychology. Maccaby, E. E., Newcomb, T. H.,and Hartley, E. L., eds., New York: Holt, Rinehart, and Winston, Inc., 1958. Lindzey, Gardner and Borgatta, E. F. "Sociometric Measure," Handbook of Social Psychology. I.Gardner Lindzey, editor. Reading, Mass.: Addison Wesley Publishing Co., 1959. Loomis, Charles P. and Pepinsky, Harold B. "Sociometry, 1937-1947: ’ Theory and Methods," Sociometry, XI (August, 1948), 262-287. MacIver, Robert and Page, Charles. Society: An Introductory Analysis. New York: Holt, Rinehart and Winston, 1949. Martindale, Donald. American Society. Princeton, New Jersey: D° Van Nostrand Company, Inc., 1960. Mills, C. Wright. The Sociological Imagination. New York: Grove Press, Inc., 1961. Moreno, J. L. "Sociometry, Comtism and Marxism," Sociometry, VIII (1945), 117-119. Moreno, J. L. Sociometry Reader. Glencoe, 111.: The Free Press, 1960. Northway, Mary. "A Plan for Sociometric Studies in a Longitudinal Programmeof Research in Child Deve10pment," Sociometry, XVII (1954), 272-281. Northway, Mary L. A Primer of Sociometry. Toronto: University of Toronto Press, 1952. Olmsted, Donald. Social Groups, Roles, and Leadership. East Lansing, Michigan: Institute for Community Development and Services, Michigan State University, 1961. Olmstead, Michael S. The Small Group. New York: Random House, 1959. 226 Parsons, Talcott. "Age and Sex in the Social Structure of the United States," American Sociological Review, VII (1942))604-611. Parsons, Talcott. Social Structure and Personality. Glencoe, 111.: The Free Press, 1964. Parsons, Talcott, Bales, Robert F., and Shils, Edward. WCrking Papers in the Theory of Action. Glencoe, 111.: The Free Press, 1953. Pepinsky, Pauline M. "The Meaning of Validity and Reliability," Educational and Psychological Measurements, IX (Spring, 1949), 39-49. Schachter, S., Ellertson, N., MCBride, D., and Gregory, C. "An Exper- imental Study of Cohesiveness and Productivity,” Group Dynamics, ResearchL and Theory. Cartwright, D. and Zander, A., eds., Evanston,Ill.: Row, Peterden and Co., 1960. Selltiz, Claire, Jahoda, Marie, Deutsch, Morton, and Cook, Stuart W. Research Methods in Social Relations. New York: Henry Holt and and Company, Inc., 1960. Shaw, Marvin, Rothschild, Gerard, and Strickland, John. "Decision Processes in Communication Nets Upon Organization and Performance in Task Oriented Groups," Management Science, I (1954), 233-250. Siegel, Sidney. Nonparametric Statistics. New York: McGraw-Hill, 1956. Simmel, Georg. The Sociology of GeorgSimmel. Trans. and Ed. Wolff, K. H., Glencoe, 111.: The Free Press, 1950. Smith, Mapheus. "Some Factors in Friendship Selections of High School Students," Sociometry, VII (1944), 303-313. Stone, Gregory P. and Form, W. H. "Instabilities in Status: The Prob- lem of Hierarchy in the Community Study of Status Arrangements," Principles of Sociology, Young, Kimball and Mack, Raymond, eds., New York: American Book Co., 1965. Rainwater, Lee, Coleman, Richard P., and Handel, Gerald. Workingman's Wife. New York: MacFadden Books, 1962. Warner, Lloyd. American Life, Dream and Reality. Chicago, 111.: The University of Chicago Press, 1962. Warner, Lloyd. Meeker, Marcia,and Eells, Kenneth. Social Class in America. Chicago: Science Research Associates, Inc., 1949. Wass, Betty. "Clothing as Related to Role Behavior of Ninth Grade Girls," Unpublished Master's thesis, Michigan State University, 1962. 227 Wax, Murray. "Themes in Cosmetics and Grooming," American Journal of Sociology, LXII (1957), 588-593. Wert, J. E., Neidt, C. C., and Ahmann, J. S. Statistical Methods in Educational and Psychological Research. New York: Appleton- Century-Crofts, Inc., 1954. Williams, Madelyn. "Opinions on Clothing, Appearance and Group Ac— ceptance as Factors in Group Cohesion of Ninth Grade Girls," Unpublished Master's thesis, Michigan State University, 1963. TE UNIV. L BRRRIE MICHIGAN STA mWIWllmllllllllmlI 3129310 4-1 111111115 4| 7 Ill“ 13